strategic positioning of international agricultural ...ageconsearch.umn.edu/bitstream/152160/2/cp...
TRANSCRIPT
Contributed paper prepared for presentation at the
57th AARES Annual Conference, Sydney,
New South Wales, 5th -8th February, 2013
Strategic positioning of international agricultural
research centres: Comparative advantage and trade-
offs from a transaction cost economics perspective
Josey Kamanda1 (Corresponding Author), Regina Birner
2 and Cynthia
Bantilan3
1Division of Social and Institutional Change in Agricultural Development
(490c) and Food Security Centre (FSC), University of Hohenheim,
Wollgrasweg 43, 70599 Stuttgart, Germany, Phone: +4971145922514,
Email: [email protected]
2Division of Social and Institutional Change in Agricultural Development
(490c), University of Hohenheim, Wollgrasweg 43, 70599 Stuttgart,
Germany
3Research Program on Markets, Institutions and Policies (MIP),
International Crops Research Institute for the Semi-Arid Tropics
(ICRISAT), Patancheru 502324, Andhra Pradesh, India
ii
© Copyright 2013 by Authors’ names.
All rights reserved. Readers may make verbatim copies of this document for non-commercial
purposes by any means, provided that this copyright notice appears on all such copies.
1
Strategic positioning of international agricultural research
centers: Comparative advantage and trade-offs from a
transaction cost economics perspective
Josey Kamanda
1 (Corresponding Author), Regina Birner
2 and Cynthia Bantilan
3
1Division of Social and Institutional Change in Agricultural Development (490c) and Food
Security Centre (FSC), University of Hohenheim, Wollgrasweg 43, 70599 Stuttgart, Germany,
Phone: +4971145922514, Email: [email protected]
2Division of Social and Institutional Change in Agricultural Development (490c), University
of Hohenheim, Wollgrasweg 43, 70599 Stuttgart, Germany
3Research Program on Markets, Institutions and Policies (MIP), International Crops Research
Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502324, Andhra Pradesh, India
Abstract
International agricultural research centres (IARCs) have a mission to reduce poverty, improve
food security, human health and nutrition, and ensure sustainable management of natural
resources. Their role in the research for development (R4D) continuum has long been a
subject of discussion, often with emphasis that they should conduct research that produces
international public goods (IPGs). However, national agricultural research systems (NARS) in
many developing countries have insufficient capacity to translate these products into welfare
benefits. This coupled with higher dependence on bilateral donors that exert pressure to show
impacts have increasingly driven IARCs to engage in participatory downstream work. This
shift has been criticized for placing emphasis on local development agendas at the expense of
IPG delivery. This paper uses insights from the literature to discuss the rationale for setting up
IARCs under the consultative group on international agricultural research (CGIAR), their
governance and transformation over the years and the critical question of how the centres
should position themselves. A conceptual framework based on transaction cost economics
and fiscal federalism literature is used to complement discussions on their comparative
advantage from a normative point of view. While low transaction intensity, asset specificity,
economies of scale and potential for spillovers are important attributes of transactions that
increase the comparative advantage of IARCs over other actors in the R4D spectrum,
contextual factors in different locations may drive centres to deviate from conducting
activities that they are best at.
Keywords: Agricultural innovation; comparative advantage; research spillovers; transaction
costs; CGIAR
2
Introduction
The World Development Report, 2008 on “Agriculture for Development” (World Bank 2007)
stresses the importance of agriculture-led growth to reduce poverty and food insecurity.
Productivity improvements, closely linked to investments in agricultural research, are key
drivers for this growth (Alston et al., 2000). Even though capacities for agricultural R&D are
low in many developing countries, farmers still largely rely on the public sector for
technology transfer (Pineiro 2007). The international agricultural research centre (IARC)
evolved as a model to improve the lives of poor rural people by increasing the productivity of
developing country agriculture (Herdt, 2012). Over time, the mandate of IARCs in the
Consultative Group on International Agricultural Research (CGIAR1
) has expanded to
include reduction of rural poverty, increasing food security, improvement of human health
and nutrition, and ensuring more sustainable management of natural resources.
Agricultural R&D follows a path from research to dissemination to uptake and impact. The
technology and knowledge generation programmes are located on the “upstream” side, while
the “downstream” side comprises delivery programmes (Kassam, 2003). There is concern that
involvement by CGIAR centres in downstream activities (D-end of the spectrum) may
directly compete with other actors and undermine incentives for building national systems
(CGIAR Science Council, 2009). The general consensus has been that the CGIAR focuses on
conducting research that produces international public goods (IPGs). However, a functional
R-D pathway does not often exist, and the link between the outputs they produce with
complementary activities that are the primary responsibility of national and local entities is
often not explicitly defined (Sagasti and Timmer 2008). The lack of adoption is attributed to
the institutional context, especially the failure of government to provide the enabling
conditions. National systems, on the other hand, may hold the impression that centre
researchers are in an “ivory tower” and not in touch with ground-level realities.
Sagasti and Timmer (2008) identify the zone of control and zone of influence by the IARCs
in the IPG delivery system. Centers may indirectly be held responsible for exerting influence
on the network of institutions and building capacity to ensure expected benefits materialize.
This supports the view that involvement in some complementary activities including
adaptation, dissemination, extension, technical assistance, policy advice, and training is
required (Pingali and Kelley, 2007). The centres have had to conduct applied and adaptive
research when developing countries lacked their own capacity to do so (Gardiner and
Chapman 2006). Declines in core funding have also led to increased dependence on bilateral
projects with donor pressure to show impact pushing the centres to conduct more
development-oriented activities. The IARCs therefore face challenges in priority setting,
targeting and strategic positioning. There is need for a clear direction in pursuing their long-
term strategic goals, while being responsive to change and demonstrating accomplishment of
1 It should be noted that although the discussion focuses on the CGIAR, the international agricultural research
system also includes other centers such as the World Vegetable Centre (formerly AVRDC), International Centre of Insect Physiology and Pest Ecology (ICIPE), International Fertilizer Development Center (IFDC), International Centre for Integrated Mountain Development (ICIMOD), Centre for Agricultural Bioscience International (CABI), Australian Centre for International Agricultural Research (ACIAR) among others, which are not part of the CGIAR.
3
short-term donor-driven objectives. They also need to address international development
concerns while responding sensitively to wishes of a broad array of local stakeholder groups
(Horton and Mackay, 2003).
This paper looks at the setup of the CGIAR system and its transformation to address
governance challenges in agricultural research, which is an issue that has often been neglected.
The outstanding debate on the comparative advantage of the centres and tension between
focus on production of IPGs versus more location-specific activities is revisited. From a
normative point of view, this is a decision on what governance structure fits different
transaction types and what level of decentralization would be ideal for each activity along the
results chain. A conceptual framework is developed, based on fiscal federalism literature
(Oates, 1972) and transaction cost economics (Williamson, 1991; Birner and Wittmer, 2004),
to analyze this trade-off.
The fiscal federalism literature has been applied in analysis of decentralization in rural service
provision (Bardhan, 2002; Birner and von Braun, 2009). The argument can be extended to
what activities should be carried out by IARCs and what should be delegated to other actors
such as national agricultural research systems (NARS) i.e. international versus national
provision of public goods. In the transaction cost economics approach, transactions that differ
in their attributes are aligned with governance structures that differ in their costs and
competence. We discuss how the framework can be applied by developing and testing
hypotheses on the types of attributes of transactions that determine how far downstream the
CGIAR centres should go.
The contribution of outputs from agricultural research to improve welfare of the poor depends
on contextual factors where the people live. Along these lines, the paper discusses how
centres may be driven to conduct activities for which they do not have a comparative
advantage. We also discuss how the institutional environment influences the probability of
success in applied and adaptive research and as well as adoption parameters with implications
on total return to investment. The review and analytical framework suggested contributes to
strategic discussions on how impact from international agricultural research can be achieved
most cost effectively.
1. Overview of the CGIAR
The world faces challenges to reduce poverty, improve food and nutrition security and
achieve sustainable management of natural resources with increasing population and threats
such as climate change. IARCs have a role to play in addressing these challenges by driving
scientific developments relevant to agriculture. International agricultural research traces its
origins to the work of the Rockefeller and Ford Foundations in the 1940s and 1950s that saw
the establishment of overseas rural and agricultural development activities (Herdt, 2012). In
1943, a pilot program2 in Mexico sponsored by the Mexican government and the Rockefeller
Foundation developed into an innovative, sustained collaboration between local and
international researchers (Ozgediz, 2012). Since most developing countries faced acute food
2 Scientists, led by Norman Borlaug, produced new higher-yielding, short-statured wheat varieties by
incorporating semi-dwarfing genes originating in Japan.
4
shortages and struggled to feed rapidly expanding populations (Zeigler and Mohanty, 2010),
the foundations invented the IARC model to exploit the emerging scientific advances to
improve the lives of the poor.
The first two centres established were the International Rice Research Institute (IRRI) in the
Philippines in 1960 and the International Maize and Wheat Improvement Center (CIMMYT3)
in Mexico in 1966. Under the leadership of champions such as Robert McNamara, a donor
support group was formed with the Food and Agriculture Organization (FAO) and the United
Nations Development Programme (UNDP) joining the World Bank as co-sponsors (Ozgediz,
2012). In 1971, international agricultural research became institutionalized in the form of the
CGIAR system, which then expanded to 15 centres (Figure 1) through the 80s and 90s.
Challenge programs (CPs) and system-wide programs were formed to shift financing
arrangements from the centres to global and strengthen partnerships among the centres, with
3 CIMMYT derives from the Spanish version “Centro Internacional de Mejoramiento de Maíz y Trigo
Alliance Office
CGIAR Members
Global
Forum on Ag.
Research
Regional
Federations &
FORA (APAARI,
AARINENA, LAC
Form on AR)
Other non-
CGIAR
Institutions
CGIAR
Secretariat
SC
Secretariat
CAS-IP Internal Audit
Gender and
Diversity Program
Chief Information
Office- ICT - KM
Countries:
Developing (25)
Developed (22)
International
Organizations (13)
Foundations (5)
IFAD
UNDP
WB
FAO
Consultative
Group
Executive
Council ExCO
AGMs
Challenge Programs
Advisory Committee
Science Council
Standing Panels
GRPS: Genetic
Resources Policy
Committee
SPIA: Impact
Assessment
SPSP: Strategies
and Priorities
SPME: Monitoring
and Evaluation
SPMS: Mobilizing
Science
Partnership Committee
PSC:
Private
Sector
CSOs
Centers
CIAT
CIMMYT
ICRISA
T
IFPRI
WARD
A
IRR
I
CIFO
R
ILRI
IITA
IWMI CIP ICARDA
Bioversity Int. WAF WFC
Communities of
Practices
Marketing Group
Executive Committee
Alliance
Board
Alliance Executive
Alliance
Subcommittees & AE
Task Forces
System Office
Figure 1: CGIAR system structure before the reform process. Source - Le Page, 2011
5
NARS and other actors. An executive council (ExCo) facilitated the work of the CGIAR with
the Technical Advisory Committee (TAC), later transformed into the science council, setting
priorities and allocating resources (CGIAR Technical Advisory Committee, 2000).
Committees were also included to provide perspectives from NGOs and the private sector.
The Impact Assessment and Evaluation Group (IAEG) continued to operate as the Standing
Panel for Impact Assessment (SPIA).
Meanwhile, the centres faced funding shortages in the 1990’s forcing them to take actions
such as downsizing of staff. There was also an increasing dependence on bilateral projects
with fears that much of the restricted funding involved project activities previously outside
the research agenda of the system.
1.1. The CGIAR Reform Process
In order to effectively harness strengths and assets of different CGIAR centres and improve
the organizational structure of the system, a reform process was initiated in 2009. It involved
adoption of an agricultural-research-for-development (AR4D) approach to take full advantage
of talents and opportunities of different actors in the wider agricultural innovation system
(CGIAR SRF, 2011). The research priorities would be guided by their potential contribution
to system outcomes in line with the CGIAR mission.
The CGIAR Consortium now binds the work of 15 centres under the CGIAR Research
Programs (CRPs) and provides a single contact point for donors. To ensure strengthened and
coordinated funding linked to system agenda and priorities, a new multi-donor funding
mechanism, the CGIAR Fund, finances the research guided by the strategy and results
framework (SRF). Independent advice is provided by a panel of leading scientific experts who
form the Independent Science and Partnership Council (ISPC). The CGIAR Fund was
designed to finance the approved CRPs through two funding windows, one for unrestricted
contributions to be allocated to CRPs by the Fund Council, and the other for contributions
where donors target specific CRPs (CGIAR SRF, 2011). However, donors preferred to have a
third window during the transitional period to provide the option of channelling contributions
directly to specific centres through the CGIAR Fund.
Figure 2: CGIAR new system structure after reform. Source - Le Page, 2011
CONSORTIUM
FUND
CONSORTIUM BOARD
Consortium CEO & Office
Centers
Partners and
Stakeholders
FUND COUNCIL
CGIAR
Research
Programs
Fund Office
Common
Services
FUNDERS
FORUM
Independent Science
and Partnership
Council
Strategy and
Results Framework
Independent
Evaluation
Arrangement
6
The new CGIAR model has faced uncertainty on how donors would respond to the proposed
harmonization, and the bureaucratization and formalism that comes with the results-based
contractual relationships for CRPs between the Fund, Consortium, lead centre, and partner
institutions. There is also uncertainty on how the relationship between the Consortium and the
centres will evolve over time, especially regarding oversight and accountability4 (Ozgediz,
2012). Hartmann (2009) sees the CGIAR reforms as taking research decisions too far away
from centre scientists who interact more frequently with national colleagues, farmers and
national governments and therefore understand local needs. In order to address the challenges
that prompted the reform process, it will be vital to develop a coordinated system of
decentralized experimentation with centralized learning (Ekboir, 2009). The role of the
centres viz-a-viz other actors needs to be clearly defined based on their comparative
advantage.
2. Comparative Advantage of CGIAR centers
In order to effectively contribute to addressing development problems, the CGIAR centres
need to justify how far down the R-D path they should go, and who will be responsible for the
next steps.
2.1. Rationale for International Agricultural Research
The first millennium development goal targets to halve the proportion of people who suffer
from extreme poverty and hunger. The 2008 World Development Report (World Bank, 2007)
stresses the importance of agriculture-led growth to achieve these targets. Although there are
differences across regions, productivity growth has driven agriculture’s global success. This
has been closely linked to investments in agricultural R&D (Alston et al., 2000; Pardey et al.,
2006; Raitzer and Kelley, 2008; Renkow and Byerlee, 2010). Estimates, of nearly 700 rates of
return on R&D and extension investments in the developing world average 43 percent a year
(Alston et al., 2000). Since NARS in many developing countries lack the R&D capacity to
meet the need for improved technologies, international agricultural research developed to
address this gap with the CGIAR being institutionalized as a key player. The types of outputs
produced by CGIAR centres are classified in Table 1.
Table 1: Classification of the types of outputs from CGIAR centres
Types of output Examples
Technological
Embodied Germplasm, parental lines for hybrids, intermediate and finished
crop varieties, farm tools and equipment
Disembodied Soil and water management techniques, crop management
practices, laboratory methods and protocols
Knowledge
base
Managerial, institutions
and policies
Participatory approaches to plant breeding or water management,
options to reduce transaction costs in input & output markets
Databases Genomic information of crops, simulation models, panel data on
rural households, commodity situation and outlook reports
Source: Authors
4 While the Consortium controls the flow of funds from the CGIAR Fund to the centers, it has limited authority
over the centers since it is their own creation.
7
They comprise technologies, either embodied or disembodied, and knowledge base in the
form of institutional options or databases. Ekboir (2009) categorizes the products as codified
(e.g. a paper or blueprint), embedded (e.g. an improved variety) or tacit (e.g. why an
experiment failed).
The traditional primary domains of advanced research institutes (ARIs), IARCs, NARS, non-
governmental organizations (NGOs) and farmers are presented in figure 3 (Craswell and de
Vries’, 2001; cited in CGIAR, 2006). Four types5 of research are identified viz. basic,
strategic, applied and adaptive research.
The CGIAR is expected to conduct strategic and applied research while working in
partnership with ARIs in basic research, and NARS in different countries in adaptive research
to diffuse the new knowledge and adjust technologies to fit relevant ecological and production
conditions across the globe (CGIAR Science Council, 2006).
Given the global mandate of the CGIAR, the IPG concept has long been discussed and
emphasized as a criterion in setting system priorities (CGIAR Science Council, 2005) to
ensure public investment in agricultural R&D obtains maximum spillovers without crowding
out national players.
2.2. The concept of International Public Goods (IPGs) in the CGIAR
Samuelson’s (1954) pure theory of public expenditure defines the concept of public goods as
used by economists. Pure public goods are differentiated from private goods by virtue of
being non-rivalrous in consumption and non-excludable. Non-excludability implies that it is
either impossible or very costly to exclude those who do not pay for the good from utilizing it,
and once the good has been produced its benefits (or harm) accrue to everyone. The non-
rivalry property means that any one person’s or consumption of the public good has no effect
on the amount of it available for others. Most public goods are impure because they exhibit
some attributes of rivalry and excludability (Kanbur 2001). Other categories of goods are thus
identified as club goods and common pool resources (Table 2).
Public goods can be defined at the local, national, regional, international or global levels6. The
view that CGIAR centres should focus on provision of IPGs began to be clearly voiced in the
5 Basic research is designed to generate new understanding, strategic research for the solution of specific
research problems, applied research to create new technologies and adaptive research to adjust the technologies to the specific needs of a particular set of environmental conditions. 6 Local public goods are available within a district, municipality or state; national public goods only within the
borders of a country; regional public goods to two or more contiguous countries within a geographic or political environment; international public goods to two or more countries across geographic, political or continental divides; and global public goods are available to all countries.
Figure 3: Primary Domains across the research continuum of INRM. Source: CGIAR Science Council, 2006.
8
late 1990s and early 2000s (Sagasti and Timmer 2008). This concept has since been a subject
of discussion in various fora (CGIAR Science Council, 2006; 2008a). The comparative
advantage of the CGIAR in producing IPGs derives partly from the fact that private firms
have limited interest in public goods since they do not have the capacity to capture much of
the benefit through proprietary claims (Pingali and Kelley, 2007). IPGs are also characterized
by coordination problems that render them unattractive to individual governments and private
agents (Spielman, 2007).
Table 2: Classification of Economic Goods
Consumption Access
Exclusive Non-Exclusive
Rival Private (eg, food, clothing, cars) Common pool (eg, air, water, soil,
landscapes, ocean fisheries)
Non-Rival
Club/Toll (eg, INTELSAT, Suez Canal, Panama
Canal, private schools, theatres, professional
associations)
Public (eg, sunshine, national
defence, lighthouses)
Source: Ryan (2006)
Harwood (2006, pp. 381) defines IPGs in the CGIAR context as:
“Research outputs of knowledge and technology generated through strategic and applied
research that are applicable and readily accessible internationally to address generic issues
and challenges consistent with CGIAR goals”.
Ryan (2006) argues that IPG characteristics are easy to define but difficult to operationalize
within the centres. IARCs face the challenge of defining what they mean by IPGs and how to
strike a balance between focusing on these versus system goals.
2.3. Criticism of the IPG Criterion
The IPG concept has been easily applied to traditional CGIAR work, like germplasm
improvement and development of new varieties, than other types of technologies or
knowledge (Table 1). Critics consider the IPG criterion as a conceptual barrier7
with
unrealistic division of labour between research and development (CGIAR Science Council
2008a). Since obstacles to achieving impact are greatest in developing countries, IPGs should
not be a shelter to hide behind the institutional bottlenecks8. The critical argument is whether
the CGIAR is focused only on producing IPGs, and not on their application. For instance, if
seed markets are a limiting factor, would producing improved lines be a 'relevant' IPG? Some9
argue that the most significant transformations led by the CGIAR took place before the
advent of “IPGs” (CGIAR Science Council 2008a). Strong emphasis on IPGs runs the risk of
intellectualizing the CGIAR mission and distancing the system from reality.
The results of research are published in scientific journals that are available internationally.
The knowledge and technology options published are often considered as IPGs even though
they may not be relevant to other countries beyond the specific laboratories, institutions and
7 Jonathan Wooley, during Special Session on IPGs at CGIAR AGM, Maputo, Mozambique November 27, 2008
8 Gebisa Ejeta, during Special Session on IPGs at the CGIAR AGM, Maputo, Mozambique November 27, 2008
9 Gebisa Ejeta, during Special Session on IPGs at the CGIAR AGM, Maputo, Mozambique November 27, 2008
9
locations where they were developed. Since journal articles may even not be accessible to
researchers in many developing countries, they can be considered as club goods as one needs
to pay to access them even though they are not rivalrous. On the other hand, working papers
available free on the internet can be considered public goods. This calls for a more specific
definition of what constitutes an IPG output.
The type of benefit to be derived from a public good and the beneficiary group that derives it
are also important considerations. The original definition of public goods did not capture the
fact that products can be publicly available but not accessible. IPGs have to be eventually put
to use by national programs, organizations or individuals in a specific location and successful
delivery will be influenced by the institutional context including policy and political systems
(Kherallah and Kirsten, 2001). Understanding of these intermediate governance and
institutional issues should be recognized in the CGIAR as process knowledge of IPG nature
that improves efficiency and likelihood of impact.
3. Governance Challenges in Agricultural Research
Governance challenges in agricultural R&D have not been adequately addressed in theoretical
and empirical literature. In this section we discuss some of them, both at the national and
international level.
3.1. Governances Challenges in National Systems
Until the 2007-2008 world food crises, public funding for agricultural research had
diminished more so in agriculture-based countries (World Bank, 2007). In sub-Saharan Africa,
although such investments increased by 20 percent in the period 2001-2008 after two decades
of almost stagnant growth, the total investment level remained low (Lynam et al., 2012). The
average rate of return (ROR) to NARS in developing countries is much lower compared to
IARCs. In Africa, the median ROR for IARCs is 83 percent higher than for NARS, while in
Asia and Pacific the gap is 72 percent indicating the need for NARS capacity strengthening
(Von Braun et al., 2008).
The heterogeneous nature of smallholder agriculture, where a large number of small farmers
are dispersed across the country (Binswanger and Rosenzweig, 1986) presents challenges.
Higher transactions costs are incurred to access them and it becomes difficult to standardize
the extension package. Research and extension workers must exercise discretion (Pritchett &
Woolcock, 2004) and provide advice tailored to varying local conditions. Considering that a
lot of staff is required on a daily basis for programs such as extension throughout the country,
there is difficulty in monitoring and supervision. The fact that they are poorly compensated
means that they lack incentives to perform effectively leading to problems such as shirking.
Efforts to use community action are also prone to the risk of elite capture where benefits
accrue to the better-off and more powerful groups. Programs such as input provision that
involve large amounts of resources distributed under conditions that are difficult to monitor
are prone to leakage and procurement challenges (Birner, 2008). In addition, although the
situation is improving, many developing countries especially in Africa have been faced with
problems such as corruption, political instability and civil war.
10
3.2. Governances Challenges in International Agricultural Research
The CGIAR, being a large and complex multilayered institution comprising a diverse group
of political (donors), operational (centres and their research partners), and strategic (advisory
bodies) players (Kassam, 2006), lends itself to governance and co-ordination challenges.
Alston et al. (1998b) observe that “market, political, and institutional failure is pervasive
since cooperative solutions among self-interested groups are hard to sustain when free-riding
on the efforts of others appears to be a viable alternative”.
In the first quarter century of establishment, the CGIAR centres remained a loosely-knit,
decentralized structure based on voluntary contributions with no constitution, by-laws, or
written rules of procedure except a framework of informal procedural guidelines (Anderson,
1998). The centres have remained independently governed with a self-perpetuating board of
trustees (Herdt, 2012). The research policies and programs were directed by centre boards and
management. The result has been a lack of system-wide vision and strategy for impact, little
sense of overall ownership, duplicate mandates and loss of system efficiency, and complex
and cumbersome governance and lack of accountability (CGIAR Independent Review Panel,
2008). The incentive structure for scientists is also an important factor since performance and
career progression is pegged in publications in peer-reviewed journals and not on ground-
level efforts to get technologies adopted.
In the early years, centres received a large share of unrestricted funds. Over time, with lack of
coordination among investors and an increasing amount of “special project” funds, the
freedom they had in pursuing long-term priorities has reduced. Some reviews of centre
activities suggest that they continue focusing on developing IPGs from all their activities. For
example, the sixth External Program and Management Review (EPMR) of the International
Crops Research Institute for the Semi-Arid Tropics (ICRISAT) recommended that “GT-
IMPI10
(should) work on the development of hypotheses that determine the IPG potential of
ICRISAT’s downstream work on technology development, testing and adaptation” (CGIAR
Science Council, 2009). While this was an important step in appraising the international
impact of ICRISAT work, the responsibility was placed upon the socioeconomics program to
reflect on all centre activities rather than the concerned scientists from each research program.
Zeigler and Mohanty (2010) also suggest that centres should focus on producing global public
goods regardless of reduction in unrestricted funding. However, as discussed in section 2.3,
certain types of products that the CGIAR centres generate from downstream involvement,
such as policies and institutional options, may be important for agriculture-led development
but difficult to judge on the IPG dimension.
Some factors present a scope for CGIAR research agenda to be indirectly influenced by donor
countries and other organizations. First, investment patterns still reflect the dominant position
and contributions of a small group of donors (Table 3). Secondly, the UN bodies, aside from
providing financial resources that support the CGIAR science advisory body, also nominate
the members and chairperson of the ISPC for approval by the CGIAR (Herdt, 2012). Third,
legal constraints under which US grant-making foundations operate do not give them a free
10
Global Theme on Institutions, Markets, Policies and Impact (GT-IMPI) later renamed as Research program on Markets, Institutions and Policies (RP-MIP)
11
hand in directing their funds. US Internal Revenue Service (IRS) regulations instituted by the
US Congress in the tax reform act of 1969 restricted the latitude foundations had in funding
and creating new organizations (Council on Foundations, 2011). Critics also see strategic
alliances with large private sector players as a trend within the CGIAR to favour industrial
agriculture and a top-down, one-size-fits-all approach to agricultural research, which ignores
the knowledge and experience of farmers, farming communities, and indigenous people
(Sharma, 2004).
Table 3: Top Donors by Decade (Amount in US$ million)
1971-1979 1980-1989 1990-1999 2000-2010
United States 105.7 United States 412.7 World Bank 426.8 United States 650.4
World Bank 42.9 World Bank 236.0 United States 392.3 World Bank 539.9
Canada 39.3 Japan 127.9 Japan 321.9 United Kingdom 389.4
Germany 33.9 Canada 103.0 European
Commission
159.3 European
Commission
337.5
IADB* 29.2 IADB* 88.8 Switzerland 149.7 Canada 298.2
United Kingdom 23.7 Germany 87.5 Germany 146.7 BMGF** 218.6
Rockefeller
Foundation
21.2 United
Kingdom
78.1 Canada 143.6 Switzerland 198.5
Ford Foundation 20.3 UNDP 72.1 Netherlands 110.3 Netherlands 185.6
UNDP 19.3 European
Commission
67.3 United Kingdom 109.7 Japan 184.0
Sweden 15.3 Switzerland 58.5 Denmark 102.8 Germany 170.6
Italy 58.5
*Inter-American Development Bank, **Bill and Melinda Gates Foundation (Began contributing in 2004)
Source: CGIAR Fund Office, 2011.
Until the Gates Foundation came into the picture, the relative importance of private
foundations and support from national governments to the CGIAR system had weakened
(Pingali and Kelley, 2007). Private research organizations have provided minimal financial
support for a system from which they also benefit (Alston et al., 1998b). This scarcity of
funds and the need to show impact have further pushed centres down the R-D continuum
(Bertram 2006). Katyal and Mruthyunjaya (2003) observed that IARCs were underfunded and
overstretched with centres being pulled downstream and compelled to oblige to pet projects of
donors.
In 2008, a comprehensive review of the structure and activities of the CGIAR was carried out
(CGIAR Independent Review Panel, 2008). The review identified the proliferation of CGIAR
entities and programs, dispersal of research focus and complexity of decision making as
severe impediments to effectiveness. The reform process described in section 1.1 was initiated
to develop a system structure that addresses these issues. The conceptual framework
presented in the next section is an approach to analyze the appropriate institutional setup for
international agricultural research if it is to achieve intended impact in the most efficient
manner. Consideration of potential governance structures is crucial especially at this time
when the CGIAR system is undergoing a reform process.
12
4. Conceptual Framework
In this section, we explore how a conceptual framework based on transaction cost economics
can be applied in analyzing cost-effective institutional design for agricultural R&D from the
global to the local using the example of the CGIAR system.
4.1. Types of Transactions in the Agricultural R-D Spectrum and Associated Costs
CGIAR centres incur direct costs that can be directly assignable to their research to produce
specific IPGs and indirect costs that cannot be directly attributed to specific research outputs.
To apply the cost-effectiveness approach, transactions or activities along the research-impact
pathway can be categorized as follows:
Planning Transactions: This includes activities like priority setting, resource mobilization
including interaction with donors and CGIAR system level duties. They are associated with
decision costs such as the direct costs of attending meetings (e.g. for strategic planning), staff
time spent in donor relations, and payments for processes, organizational units and personnel
required to maintain the CGIAR as a system. The interface costs are higher with reduced core
funding as centres have to deal with a broader set of bilateral donors with different interests
and requiring different mobilization strategies. Cutting down on these interfacing costs will
lead to financial savings, but there is the opportunity cost regarding the quality of decisions
made. This is a trade-off that needs to be captured as making suboptimal decisions may lead
to decision-failure costs (Birner and von Braun, 2009).
Production Transactions: This includes activities such as setting up the required
infrastructure, human resources and partnerships required in order for the centres to do their
business (research), as well as the actual conducting of research. It includes set-up and
maintenance costs, research costs and costs of shared services such as research support,
financial management, procurement, personnel management project management, and
information technology. It also includes costs of maintaining and distributing CGIAR crop
genetic resources which requires sustained funding. Production failure costs may be incurred
if the research is delegated to a partner that does not have sufficient capacity.
Promotion Transactions: Subject to availability of funding and manpower, promising
technologies that have been produced and tested are promoted and disseminated to potential
beneficiaries. The process includes costs of extension, technical assistance, policy advocacy,
and training and capacity building for partners. It also includes the costs of shared services.
Monitoring and Evaluation and Reporting Transactions: Impact assessment and project
reporting activities involve costs for data collection, analysis and write-up. Again these costs
escalate when the centres have a large number of bilateral projects with small budgets that
need to be reported separately. Projects that do not budget for evaluation activities may be
unable to show accountability to donors and therefore run the risk of losing additional funding.
Transactions at the User Level: These are the activities that beneficiaries (e.g. farmers) need
to undertake to access the technologies/ knowledge. They incur some costs aside from the
cost of the products e.g. time and money used for travel or to access an extension agent.
13
The reform process of the CGIAR has implications on the above costs. Overall interface
activities of the centres are expected to diminish as the consortium serves as the main
interface with donors. The majority of research will be organized under the CRPs with an
increasing proportion of funds being mobilized and flowing through the CGIAR Fund. This
means that there will be a common monitoring and evaluation framework and reduced
number of bilateral projects that have to be reported separately. However, there might be
additional system level costs and a loss of benefits associated with direct engagement between
centres and donors.
4.2. Comparative Cost-Effectiveness of Different Governance Structures
In order to achieve an economizing result and higher total welfare from a given set of
resources, each transaction should be assigned to the actor who, in relative terms, is best at
carrying it out i.e. has a ‘comparative advantage’. For a start, we consider the above
transactions as activities being carried out by either an IARC or a NARS. According to the
discriminating alignment hypothesis, we hypothesize that low transaction intensity, asset
specificity, economies of scale and potential for spillovers are important attributes of
transactions that increase the comparative advantage (cost-effectiveness) of IARCs over
NARS in carrying out the transaction.
TCc
Participatory-Adaptive Applied Strategic Basic
TCh
$ TC
n
a
International Agricultural
Research Centers
(IARCs)
National Agricultural
Research Systems
(NARS)
Increased NARS
capacity
Transaction
costs + other
costs arising for
achieving a
specific
impact
Attributes
* Low transaction intensity
* Asset Specificity
* Economies of Scale
* Potential for spillovers
a1
TCi
0 a2
Hybrid
Structure
Figure 4: Comparative cost-effectiveness of conducting research by IARCs versus NARS. Source: Based on
Williamson (1991) and Birner and Wittmer (2004)
Hypothetical cost curves are shown in figure 4 when the above transactions are carried out by
an IARC (TCi) versus a NARS (TC
n). This is analogous to comparing a more centralized
(IARC) to a more decentralized (NARS) structure. The vertical axis indicates costs arising for
carrying out the activity; horizontal axis combination of attributes which increase the
comparative advantage of IARCs in carrying out the activity. As these attributes become more
important, IARC transaction costs increase less rapidly. This is indicated by the reduced slope
of the respective cost curve. If these attributes are not relevant (moving to the left-hand side
on the horizontal axis), NARS have a comparative advantage over IARCs. From point a1
14
onwards IARCs have a comparative advantage over NARS for performing the respective
transaction. If capacity of NARS is increased, it will be able to carry out the activity at lower
costs, indicating a downward shift of the respective cost curve (TCc). The point from which
IARCs have a comparative advantage over NARS for thus shifts from point a1 to a2.
While the above discussion considers the comparison between IARCs and NARS, we
recognize that there are many other actors in the agricultural R&D process. Most often,
IARCs work in collaboration with these partners on joint research projects with each partner
doing those types of transactions for which it has a comparative advantage. This type of
hybrid governance structure is represented in figure 4 by the cost curve TCh. This should give
higher total welfare for a given set of resources as each actor is assigned what they can carry
out most efficiently. Linking back with the discussion in section 2, the transactions under
basic and strategic research (right side of horizontal axis) have attributes of low transaction
intensity, asset specificity, economies of scale and higher potential for spillovers compared to
applied and participatory research (left side of horizontal axis).
4.3. Effect of Contextual Factors
The framework presented so far discusses what the IARCs and other actors should ideally do
along the R-D process from a normative point of view. In reality, generation and diffusion
patterns for innovations depends on complicated set of factors and stakeholder interactions.
There is a need to understand how this complexity affects the effectiveness of agricultural
R&D and the prospects of reaching the poor. The innovation systems perspective (World
Bank, 2006) promotes systems thinking in agriculture through careful coordination among
many stakeholders. The national agricultural innovation system in a country comprises the
agricultural research and education system, the interactions of bridging institutions from the
public, private and civil society sectors for technology delivery, and the ultimate users within
the agricultural value chain such as farmers, traders, and input suppliers. The system is also
influenced to a great extent by government policies related to agriculture such as property
rights, taxation and trade policies, investment in infrastructure and extension as well as the
informal socio-economic and cultural landscape.
The relationship between institutions and economic development has been well studied
(North, 1989; Nelson and Sampat, 2001; Acemoglu et. al, 2005; Butkiewicz and Yanikkaya,
2006; Keun and Byung-Yeon, 2009). In the context of agriculture, Monchi and Meng-Chun
(2008) use the six World Bank aggregate governance indicators11
to examine whether the
differences across-countries have an effect on agricultural productivity. They conclude that
given the same amounts of agricultural inputs, the same education level, and the same climate
condition, a country with better governance can generate more agricultural outputs.
Different localities may therefore show different adoption and diffusion patterns because of
differing social, cultural, institutional and political environments. For instance, the green
revolution in Asia is often seen as technological revolution. However, the success had a lot to
do with political interest of India to become food sufficient as well as willingness of the US
11
World Bank aggregate governance indicators include rule of law, control of corruption, government effectiveness, regulatory quality, voice and accountability, and political stability.
15
government and donors such as the World Bank, and Rockefeller and Ford Foundations to
provide support. C. Subramanian, the then Indian Minister for Agriculture, championed
institutional changes in the agricultural innovation system of India that enabled the green
revolution to materialize (Bhagat, 1998). Along the same lines, while the CGIAR has focused
on advancing agricultural science and its application to productivity growth, actual realization
of its mission depends on the institutional and policy context in target countries. There are
huge variations across locations and across different commodities making it difficult to apply
similar intervention strategies in different regions.
A CGIAR Science Council (2008b) report on ethical challenges identifies the responsibility
of centres depending on the context in the target country. In cases where NARS lack capacity,
they should be empowered through training programs or by performing research in
partnership them. If there are NARS with capacity12
but lack the will to engage in the problem,
the CGIAR can engage in advocacy. In some regions there might be serious constraints
simply because no other agents appear to be available. It would then be an ethical goal that
the CGIAR engages in the problem. In the process, the CGIAR should keep an eye for the
general aspects of the problem in order to build up knowledge that may be useful in other
regions.
The case where NARS are too inefficient that CGIAR centres always have a comparative
advantage is illustrated in figure 5. This would imply that the respective cost curves do not
intersect at any point regardless of how important the relevant attributes become.
$ TC
n
a
International Agricultural
Research Centers
(IARCs)
Inefficient National
Agricultural Research
Systems (NARS)
Transaction
costs + other
costs arising for
achieving a
specific
impact
Attributes
* Low transaction intensity
* Asset Specificity
* Economies of Scale
* Potential for spillovers
TCi
0
Figure 5: Comparative cost-effectiveness of conducting research by IARCs versus NARS when the later are
too inefficient. Source: Based on Williamson (1991) and Birner and Wittmer (2004)
12
Over the years, R&D capacity and institutional structures have indeed developed in some countries such as India and China, resulting in much lower costs relative to possible impacts from research.
16
Research priorities of the IARCs may also be the outcome of political decisions at the system
or centre level. Some of the conditions that might drive centres to conduct activities for which
they do not have a comparative advantage have been discussed in section 3.2. In the next
section we further look at relevant attributes and contextual factors, based on insights from
literature, and their influence on how the centres position themselves in R-D spectrum.
5. Insights from the Literature
5.1. Transaction Costs Economics
The transaction cost economics approach applies the discriminating alignment hypothesis
where transactions that differ in their attributes are aligned with governance structures that
differ in costs and competence (Williamson, 1991; Birner and Wittmer, 2004). It seeks to
identify the most cost-effective governance structures or institutional set-up for achievement
of a given outcome.
Transaction intensity has been used to characterize transactions in service delivery (Pritchett
& Woolcock, 2004; Birner and Linacre, 2008; Birner and von Braun, 2009). This attribute
indicates the (1) frequency (Williamson, 1991) and (2) spatial dispersion of transactions. For
instance, molecular breeding is less transaction intensive, compared to extension work since
the former requires fewer interactions with the farmers. We hypothesize that low transaction
intensity is an important attribute that increases the comparative advantage of IARCs over
NARS in carrying out the activity.
H1: Activities that require higher transaction intensity between the research organization and
beneficiaries can be carried out more cost-effective by delegating them from IARCs to NARS.
The capacity to exploit economies of scale in agricultural R&D at a global scale is linked to
the specialized assets the centres possess. Williamson (1991) defines asset specificity as
degree to which an asset can be redeployed to alternative uses and by alternative users
without sacrifice of productive value. He identifies six types of asset specificity; site
specificity, physical asset specificity, human-asset specificity, brand name capital, dedicated
assets, temporal specificity.
CGIAR centres have specialized physical assets in the form of research facilities and
equipment that can be used to undertake more advanced and complex lines of research
compared to NARS. The specialized knowledge gained over years by the large pool of
scientists and managers with experience in international research co-ordination and
facilitation is also a human asset. Considering that each centre has a specific mandate
(livestock, certain crops, specific agro-ecoregions e.g. semi-arid tropics or specific natural
resources e.g. water), the physical and human assets are specialized and difficult to deploy for
another purpose. For example, the genebank in a commodity centre will only contain
germplasm accessions for their mandate crops that cannot serve other crops needs in terms of
seeds. This represents a “hold-up” situation except for cases where there are across-
commodity spillovers representing benefits for multiple crops from a single genebank. It can
be hypothesized that asset specificity is an important transaction attribute that increases the
comparative advantage of IARCs over NARS.
17
H2: As asset specificity increases IARCs will incur less cost than NARS by internally carrying
out transactions that require those assets.
On the other hand, agricultural research, even for a specific crop mandate, requires a
multidisciplinary approach e.g. integrated genetic and natural resource management approach
(Twomlow et al., 2008). Some form of site specificity is required where synergy across
themes is achieved when stations are located in a “cheek-by-jowl” relation to complement
each other and economize on inventory and transportation expenses (Williamson, 1991). For
instance, centres have facilities like gene bank, molecular lab, greenhouse etc. as well as
human resources comprising molecular scientists, breeders, pathologists and agronomists who
may all be working on the same crop.
CGIAR research facilities are also located in poor developing countries representing many
agro-ecosystems. This facilitates highly contextualized, problem-driven research that is more
difficult and costly to undertake from ARIs based in developed countries (Herdt, 2012). The
CGIAR comprises a group of 64 member countries and organizations (CGIAR Fund Office,
2011) committed to addressing global development challenges through international
agricultural research. Because of its large and diverse international network, the CGIAR
system could play a bridging role in innovation networks (Ekboir, 2009). Its recognition as a
key player in international agricultural research represents some sort of brand name capital.
However, circumstances such as the social unrest experienced in Syria in 2011, which
affected ICARDA operations, illustrate the downside of investments in large infrastructure
that have to be abandoned. The challenge is to maintain flexibility under such limiting
conditions while at the same time continuing with high-tech work that requires these
resources.
5.2. Fiscal federalism literature
The fiscal federalism literature (Oates, 1972) has been developed with a focus on analyzing
decentralization levels (local versus national governments) in service provision, but the theory
can be applied to other levels. A critical consideration for research managers in the CGIAR is
whether to develop centralized research programs or adopt a more decentralized structure.
They also need to decide what R&D activities should be carried out by the centres themselves
and what should be delegated to other partners.
The appropriate regional (De Janvry and Kassam, 2001; Douthwaite et al., 2005) and
institutional structure for organizing research programs is an important decision point. Alston
et al. (1998b) observe that this should vary according to the nature of the research. The
concept of aggregation of knowledge production (Spielman, 2007) refers to the way in which
the contributions of individual countries and actors determine the total quantity of knowledge
produced, or how scarce resources for research are most effectively distributed among
countries. Aggregation addresses the issue of what the research priorities should be, where the
research should be conducted, and how its benefits should be generated.
The fiscal federalism literature identifies three important attributes of transactions, which
influence the appropriate level of decentralization: economies of scale, spillovers and
externalities, and the heterogeneity of local preferences.
18
If economies of scale in carrying out a particular kind of research are large, a more centralized
system is likely to reduce the costs. Where technical knowledge is relevant, IARCs may be
more appropriate to be able to exploit economies of scale in providing or utilizing this
knowledge. It can therefore be hypothesized that economies of scale is an important
transaction attribute that increases the comparative advantage of IARCs.
H3: IARCs have a comparative advantage over NARS in conducting activities with high
economies of scale.
Where potential for replication and inter-regional transfer of research results or spillovers is
high, research programs can be more centralized. International spillovers are likely highest for
a commodity like wheat, which is grown in relatively homogeneous production environments,
with little variability in local tastes and preferences for quality characteristics (Byerlee and
Traxler, 2001). However, if preferences for the commodity differ across countries, then it
would be more appropriate to have decentralized structures to address these local needs.
5.3. The potential for spillovers
The term “spillovers” has been used in the international agricultural research community
before the 1990’s (Davis et al., 1987). Bantilan and Davis (1991) identify three types of
spillovers, namely: across-location, across- commodity and price spillovers. Technologies are
said to have spillover potential if they have applicability to other agro-ecological locations or
for a different crop (Deb and Bantilan, 2001; Shiferaw et. al, 2004). Price spillovers occur
when the technological change at a specific location increases supply of the commodity and
changes the price at other locations through trade.
Spillovers has been used to refer to the transfer of benefits from one production environment
to another, but also to the transfer of benefits from research carried out in one country to
another. Biophysical scientists easily relate to the applicability13
across agro-ecoregions since
they are interested in estimating the potential yield of technologies in production
environments different from where the research was targeted. Research managers or policy
makers are more interested in the final welfare gains and their distribution across countries or
regions to guide priority setting and resource allocation.
Some attempts have been made to quantify spillover benefits from international agricultural
research. Maredia and Byerlee (1999) quantified spillover benefits for improved wheat
germplasm across agro-ecological boundaries. Spillovers from research in one region within a
country to another have also been estimated. For example, Alston et. al (2011) measure the
returns to the United States public agricultural research with spillover benefits to across states.
Developed country agricultural research systems also benefit from the technology spillovers
generated by the CGIAR; Brennan (1986) measured the benefits to Australian wheat breeding
programs of access to breeding materials from CIMMYT. Brennan and Bantilan (2003) and
Brennan et. al. (2003) use case studies of production spillovers to Australia from the work of
ICRISAT and the International Centre for Agricultural Research in the Dryland Areas
(ICARDA) respectively. Pardey et al. (1996) measured benefits to US wheat and rice
13
This concept is used in the elicitation of spillover indices across research domains, and the resulting matrix applied to compute the total welfare gains.
19
production from germplasm developed at CIMMYT and IRRI. Although it may seem against
the philosophy under which the CGIAR was established, these assessments of spill-ins to
developed countries can help justify use of taxpayers’ money from those countries to fund
international agricultural research.
If the spillover potential of research outputs is high, research programs and infrastructure can
be centrally set up with assurance that the products can be transferred and applied in similar
environments elsewhere. It can therefore be hypothesized that potential for spillovers is an
important attribute that increases the comparative advantage of IARCs in carrying out a
particular activity compared to NARS.
H4: It is more cost-effective to have centralized IARCs conducting activities that have high
potential for spillovers rather than decentralized NARS
Depending on whether finished products, prototypes or just knowledge is transferred, Ruttan
(1975) distinguished three phases or levels of agricultural technology transfer as materials
transfer, design transfer, and capacity transfer. While technology transfer literature has
focused on material and design transfer, it is important to understand processes that lead to
capacity transfer. Learning from actor-oriented studies (Biggs, 2008) of situations where
positive socio- economic and welfare benefits have been realized is required. However, past
analyses of spillovers have mainly considered the agro-climatic characteristics despite the fact
that there are many exceptions relating varying economic, social and political institutions
(Feder et al. 1982; Deb et al., 2004). Harwood et al. (2006) discuss the international
applicability of strategic approaches to integrated natural resource management (INRM)
research. Several examples of location-specific INRM projects that generated knowledge with
spillover potential are presented to illustrate that appropriately designed research with
development components can generate different types of IPGs.
Spillover benefits from research conducted in developed countries may at times not be
available to developing countries. Pardey and Alston (2010) attribute this to the shift in rich-
country R&D agendas away from productivity gains in food staples to other aspects such as
environmental effects, food quality, and the medical, energy, and industrial uses of
agricultural commodities. They also argue that technologies that are applicable may not be as
readily accessible because of increasing intellectual property protection of privately owned
technologies and the expanding scope and enforcement of bio-safety regulations. Finally,
those technologies that are applicable and available are likely to require more substantial local
development and adaptation.
Contextual differences across countries will, therefore, shape the uptake of technologies and
subsequently the spillover benefits generated. This calls for a development of proof of
concept hypotheses to confirm the contextual factors that influence the probability of success
in development, diffusion and impact of new technologies across countries. Based on the
discussion of contextual factors in section 4.3, we hypothesize that CGIAR centres may not
apply consistent positioning strategies because of differences in donor demands as well as
huge variations in the institutional environment across locations and commodities.
20
H5: IARCs will not apply consistent positioning strategies because of differences in donor
demands and huge variations in context across locations and commodities.
5.4. Implications of Institutional Design on Returns to Investment
Economic rates of return from research products are most commonly assessed using cost-
benefit analysis. Qualitative methods can complement quantitative measurements by
providing information on the institutional attribution, uptake, and influence of the research
output under analysis (Walker et. al, 2008).
In cases where more than one institution has been involved in development and dissemination
of a technology, attribution of net benefits presents a challenge. Since the benefits will be a
joint output of the organizations involved and the associated costs, a full description of the
role played by each actor in the R&D process (Figure 6) is required. Reducing the total costs,
including transaction costs of planning, research, technology transfer, monitoring and
evaluation and uptake, through appropriate institutional design will result in higher internal
rates of return. Aside from the estimated net benefit per unit of adoption, referred to as the
unit cost reduction (Alston et al., 1998a), the institutional choice has an effect on other
parameters including the probability of success in research, capacity to conduct adaptive
research and the actual likelihood, timing and scale of adoption.
CGIAR centres have networks and respect in the regions where they work that comes from
long experience and contribution within the regions (Harwood et. al, 2006). They can
therefore play a facilitation role for a range of institutions which will influence the adoption
parameters in two ways. First is advancing adoption so that benefits materialize earlier as
indicated in Figure 6 by the reduced time to reach maximum adoption from T1 to T2. The
other is an increase in the total level of adoption from AMax
to AMax2
.
Wider Adoption
T1 T2
AMax2
Reduced
Adoption lag
Ceiling Level
15
AMax
100%
Level of
Adoption
Innovative
Research Adoption Pathway
0 4 8
Adaptive Research
/ Development
Figure 6: The Research-Adoption Pathway
Mausch et. al (2013) use the example of benefits to different countries from groundnut
research targeting the homogenous zone with highest total benefits globally to illustrate this
point. With the adoption constraint, the benefits to India are about 1,040 million USD which
21
is the real world scenario. By lifting the adoption constraint, the welfare benefits to India
increase and are equal to an ideal world scenario at about 1,600 million USD. Since adaptive
capacity is 1, we need to focus on intervention strategies that will enhance the adoption
parameter. Before this is done, the binding constraints to adoption need to be identified and
alternative solutions sought. In order to get the maximum benefits at least possible cost, the
institutions that will carry out the identified solutions most efficiently (having a comparative
advantage), should be assigned the responsibility. In many African countries, adaptive
research capacity is also low and the reasons for this can only be understood through an
institutional analysis. This will vary not only across countries but also across crops, thus the
need for disaggregated contextual case studies. Alongside breeding, efforts can then be made
to enhance both adoption and adaptive capacity in these countries. Using the example of
Malawi, lifting only the adaptive capacity constraint will increase the benefits to the country
from 14 million USD (real world scenario) to about 35 million USD. If the adoption
constraint is lifted, the benefits further jump to about 50 million USD.
Ex-ante impact assessment studies conducted to guide priority setting and targeting decisions
take into account agro-ecological conditions, such as the length of growing period (LGP14
),
with assumptions regarding the adaptive capacity and adoption parameters. These additional
parameters depend on local conditions. Although there have been attempts to characterize
these conditions using indicators such as market access (Omamo et. al, 2006) and agricultural
R&D capacity15
, data describing them are patchy and of questionable quality in many
developing countries. In addition, these aggregate indicators do not take into account
differences between commodities in a specific country.
Where resources are not available to conduct extensive surveys, the adaptive capacity and
adoption estimates are elicited as expert judgments. These are worked out implicitly in the
experts mind based on their internalized experience in the R&D cycle in the locations of
interest. However, this knowledge is likely to be lost when they leave the organization.
Studies of the R&D process and institutional drivers of uptake and impact for different crops
in different regions will make this implicit knowledge more explicit for future use. This
would reveal the considerations that underlie expert judgments, which are often not well
documented in adoption studies.
6. Applying the Framework
The suggested framework is not designed to identify the most efficient of all possible
governance structures but rather to compare which among a selected set is the most cost-
effective taking into account the contextual conditions (Williamson, 1991). Empirical
research would provide a set of alternatives whose feasibility (administrative, fiscal, and
political) can then be assessed. Based on the above review of literature, four hypotheses are
derived on the attributes of transactions for which IARCs have a comparative advantage over
NARS (Table 4) and one based on the effect of contextual factors (H5):
14
http://www.fao.org/ag/againfo/programmes/en/lead/toolbox/Refer/AgroeZon.htm 15
http://www.asti.cgiar.org/
22
H1: Activities that require higher transaction intensity between the research organization and
beneficiaries can be carried out more cost-effective by delegating them from IARCs to NARS.
H2: As asset specificity increases IARCs will incur less cost than NARS by internally
carrying out transactions that require those assets.
H3: IARCs have a comparative advantage over NARS in conducting activities with high
economies of scale.
H4: It is more cost-effective to have centralized IARCs conducting activities that have high
potential for spillovers rather than decentralized NARS.
H5: IARCs will not apply consistent positioning strategies because of differences in donor
demands and huge variations in context across locations and commodities.
Empirical research is required to test whether these hypotheses apply for specific innovations
by collecting data on different transactions in the R&D process, relevant attributes, and
contextual factors. Transactions along the impact pathway can be elicited from researchers,
partners and beneficiaries using process-influence16
maps to understand the technology
development and uptake process. Information on costs incurred by various actors also needs
to be collected.
Table 4 summarizes the attributes of transactions with an assessment, based on the literature
reviewed in section 5, on the role that each attribute plays for each of the transactions.
Transactions with high asset specificity, economies of scale and spillover potential should be
ideally assigned to a centralized institution (IARC) while those with high transaction intensity
to a more decentralized institution (NARS or other partner). While these implications are
easier to derive for production and user level transactions, the other cases involve trade-offs
depending on context and intended objective of those activities.
For planning transactions within the CGIAR, the governance and funding structure of the
system and centres determines how strategic plans are developed. The new system under
CRPs will exploit economies of scale and reduce transaction costs of interface activities, but
the opportunity cost is the risk of driving research decisions further away from local needs.
The tools and methodologies used in priority setting and targeting such as models for
forecasting, scenario analysis and ex-ante impact assessment can be applied elsewhere
representing a spillover potential. Regarding asset specificity, while specialized physical
assets may not be required in planning, the experience of scientists and partners is important.
This expertise may at times only be applicable to specific commodities.
In the promotion stage, activities like extension, capacity building and policy advocacy
require high transaction-intensity suggesting a more decentralized organization. However, this
is not straightforward. The transaction intensity, asset specificity, economies scale and
spillover potential will depend on what is being promoted. For example, as compared to
information on new varieties, guidance on crop management practices such as tillage
operations, spacing or methods of seed placement and fertilizer application requires more
interactions with the farmers and discretion (Birner and von Braun, 2009) making it difficult
16
http://netmap.wordpress.com/process-net-map/
23
to standardize. Spillover effects can still arise from location-specific activities if there is a
conscious intention to test the techniques used and draw lessons that can be adapted for
application elsewhere e.g. on extension models, policy processes etc.
Table 4: Attributes of transactions in the agricultural R&D process
Transactions
Type of
Knowledge
Required
Relevance of Attributes
Transaction
Intensity
Asset
Specificity
Economies of
Scale
Spillover
Potential
Planning Local &
Technical
Dependent on
governance &
funding structure
Medium High Medium
Production Technical Low High High High
Promotion Local &
Technical
Dependent on
promotion
objective
Dependent on
promotion
objective
Dependent on
promotion
objective
Dependent on
promotion
objective
Monitoring /
Evaluation
& Reporting
Local &
Technical
Dependent on
monitoring
objective
Dependent on
monitoring
objective
Dependent on
monitoring
objective
Dependent on
monitoring
objective
User Level Local High Low Low Low
Source: Authors, Adapted from Birner and von Braun (2009)
At the monitoring and evaluation stage, data on performance of innovations has to be
collected through interactions along the R&D chain. Depending on resource constraints, full-
fledged surveys or expert judgments that are less transaction intensive may be used. The
nature and scope of the evaluation will determine whether specialized expertise is required.
Again, tools and methods used e.g. for ex-post impact assessment can be applied elsewhere
and so can the lessons learnt from the evaluation if properly documented.
The standard approach in empirical transaction cost economics does not require a
measurement of transaction costs (Shelanski and Klein, 1995). Empirically quantifying
attributes of transactions may be challenging since variables such as asset specificity are
difficult to measure. Although some surveys have used scaling methods (Brown and Potoski,
2003), such data are subject to the general limits of survey data since that they are based on
the stated beliefs of respondents rather than those revealed through choice. The measurements,
based on ordinal rankings, are also difficult to compare across institutions (Shelanski and
Klein, 1995). The potential for spillovers can be estimated through subjective estimates of
experts, or through objective estimates based on data reflecting applicability of a new
technology across environments and using economic surplus models to quantify potential
benefits across countries (Deb and Bantilan, 2001). Comparisons can be made through the ex-
post assessment of actual impacts. Transaction intensity can be quantified e.g. by counting the
number of visits or meetings that are usually required for a certain activity like extension.
To collect data on contextual factors, it will be useful to rely on the respective literature
(Birner and von Braun, 2009). There is a wide array of literature on the dynamics of adoption
and especially the factors that influence farmer adoption decisions. However, understanding
of “what works” in diverse circumstances and the processes driving outcomes is still far from
complete. Increased attention needs to be paid to the policy and institutional context that
shapes agricultural technology development and uptake in different countries. Research is
required to understand constraints in the entire innovation process comprising technology
24
production, supply and use of different commodities. Case studies can be carried out of
research programs that have achieved the best results as well as those that have experienced
limiting political, cultural and institutional constraints to adoption in different countries. The
synthesized lessons on contextual factors can help governments and the international
development community in targeting appropriate investments and policy reforms while
bearing in mind the local political economy.
Conclusions
Agricultural research for development needs to address a wide range of issues facing
resource-poor farmers in different countries. For the intended benefits to be realized
investments at the national, regional and international level are needed but an important
decision point is who should do what. The analysis undertaken in this paper reviews the
rationale for setting up IARCs, concerns regarding governance of agricultural research, and
the ongoing CGIAR reform process to address some of these concerns. The critical question
of how CGIAR centres should position themselves in the R-D spectrum is re-examined.
Review of discussions in literature and various fora identifies a gap in addressing the dilemma
in an objective way.
The conceptual framework presented applies transaction cost economics perspectives to
explore appropriate institutional options for carrying out activities along the R-D chain. This
is a useful basis for strategic discussions on how far downstream the CGIARs should go in
order to achieve impact from agricultural R&D most cost-effectively. Based on consideration
of the relevant attributes of transactions and contextual factors, one can make trade-offs on
whether to assign an activity to IARCs, NARS or other actor in the innovation system.
Still, it is important to note that a complicated set of factors such as availability of funds and
political pressure e.g. donor preferences will influence the decision to carry out a specific
activities. Capacities of actors along the R-D spectrum differ across locations signaling the
need for a case by case analysis of the institutional environment across countries and
commodities in order to plan suitable intervention strategies. However, priority setting for
IARCs has mainly focused on targeting of research across production environments without
careful consideration of how far down the R-D continuum to be involved based on the
institutional context. Ultimately, positioning strategies and institutional design to enable
impact from CGIAR centres should be differentiated across matrices of biophysical and
institutional conditions.
Acknowledgments
We would like to acknowledge the support of the Food Security Centre at University of
Hohenheim and by extension the German Academic Exchange Service (DAAD), the German
Federal Ministry for Economic Cooperation and Development (BMZ) and Foundation Fiat
Panis for making financial resources available for this study. The funding sources had no role
in the study design, data collection, analysis and write-up, and in the decision to submit the
article for publication. We thank Dave Hoisington, Jeff Davis and Rupsha Banerjee for their
comments on earlier versions of this paper.
References
25
Acemoglu Daron, Johnson Simon, Robinson James A. 2005. Chapter 6 Institutions as a
Fundamental Cause of Long-Run Growth, In: Philippe Aghion and Steven N. Durlauf,
Editor(s), Handbook of Economic Growth, Volume 1, Part A, pp 385-472
Alston, Julian M. & Andersen, Matthew A. & James, Jennifer S. & Pardey, Philip G., 2011.
The Economic Returns to U.S. Public Agricultural Research. American Journal of
Agricultural Economics, Agricultural and Applied Economics Association, vol. 93(5), pages
1257-1277.
Alston, Julian M., Connie Chan-Kang, Michele C. Marra, Philip G. Pardey and TJ Wyatt.
2000. A meta-analysis of rates of return to agricultural R&D, Ex Pede Herculem? Research
Report 113, IFPRI, Washington DC.
Alston, J.M., G.W. Norton, and P.G. Pardey. 1998a. Science under scarcity: Principles and
practice for agricultural research evaluation and priority setting. CAB International,
Wallingford, UK.
Alston, Julian M., Pardey, Philip G. and Roseboom, Johannes. 1998b. Financing Agricultural
Research: International Investment Patterns and Policy Perspectives. World Development Vol.
26, No. 6, pp. 1057-1071, 1998
Anderson, Jock R. 1998. Selected Policy Issues in International Agricultural Research: On
Striving for International Public Goods in an Era of Donor Fatigue. World Development Vol.
26, No. 6, pp. 1149-1162, 1998
Bantilan, M.C.S., and Davis, J.S. 1991. Across-commodity spillover effects of research and
opportunity costs in a multiproduct production environment. ACIAR/ISNAR Project Paper
Series no. 30. Canberra, Australia: Australian Centre for International Agricultural Research
and The Hague: International Service for National Agricultural Research.
Bardhan, P. 2002. Decentralization of Governance and Development. Journal of Economic
Perspectives, 16 (4), 185–206.
Bertram R. 2006. Beyond International Public Goods: But Just How Far? Brief prepared for
the Workshop on “Positioning the CGIAR in the Global Research for Development
Continuum”, the Hague, Netherlands, 19 May 2006.
Bhagat Rasheeda. 1998. A visionary and a statesman. Frontline National Magazine, from the
publishers of The Hindu. Vol. 15, No. 05. March 7-20, 1998.
http://www.hindu.com/fline/fl1505/15051130.htm (Accessed 1st March 2012)
Biggs, Stephen. 2008. Learning from the Positive to Reduce Rural Poverty and Increase
Social Justice: Institutional Innovations in Agricultural and Natural Resources Research and
Development. Experimental Agriculture, Volume 44, pp. 37–60. Cambridge University Press
Binswanger, H.P and Rosenzweig, M.R. (1986). Behavioural and Material Determinants of
Production Relations in Agriculture. Journal of Development Studies Vol. 22 no.3: 503539.
Birner, R. and von Braun, J. 2009. Decentralization and Public Service Provision: A
Framework for Pro-Poor Institutional Design, in: Ahmad, Ehtisham and Giorgio Brosio
26
(Eds.): Does Decentralization Enhance Poverty Reduction and Service Delivery? Edward
Elgar, Cheltenham, UK, pp. 287-315.
Birner, R. and Linacre, N. 2008. Regional Biotechnology Regulations: Design Options and
Implications for Good Governance. IFPRI Discussion Paper 753. International Food Policy
Research Institute, Washington, DC.
Birner, Regina. 2008. Can Decentralization and Community-Based Development Reduce
Corruption in Natural Resource Management: Insights from Irrigation in Ghana and Forestry
in Indonesia, Paper presented at the 13th International Anti-Corruption Conference organized
by Transparency International. Athens, Greece October 30 – November 4, 2008.
Birner, R. and Wittmer, H. 2004. On the ‘Efficient Boundaries of the State’: The Contribution
of Transaction Costs Economics to the Analysis of Decentralization and Devolution in
Natural Resource Management, in: Environment and Planning C: Government and Policy,
Vol. 22 (5), pp. 667 – 685.
Brennan, John P, Aw-Hassan, Aden, Nordblom, Thomas L. 2003. Influence of spillovers to
Australia on impacts of the International Center for Agricultural Research in the Dry Areas,
Food Policy, Volume 28, Issues 5–6, October–December 2003, Pages 471-485.
Brennan, JP and Bantilan, MCS. 2003. Price and yield effects of spillovers in international
agricultural research: Evidence from ICRISAT and Australia. Agricultural Economics 28(2):
87-97.
Brennan, J.P. (1986). “Impact of wheat varieties from CIMMYT on Australian wheat
production”, Agricultural Economics Bulletin No. 5. Department of Agriculture New South
Wales, Sydney.
Brown, T.L. and Potoski, M. 2003. Transaction Costs and Institutional Explanations for
Government Service Production Decisions. Journal of Public Administration Research and
Theory, 13 (3), 441–68
Byerlee, D. and Traxler, G. 2001. “The role of technology spillovers and economies of size in
the efficient design of agricultural research systems”. In: Alston, J., Pardey, P., Taylor, M.
(Eds.), Agricultural Science Policy: Changing Global Agendas. Johns Hopkins University
Press, Baltimore.
Butkiewicz, James L. and Yanikkaya, Halit. 2006. Institutional quality and economic growth:
Maintenance of the rule of law or democratic institutions, or both?, Economic Modelling,
Volume 23, Issue 4, pp 648-661
CGIAR Fund Office. 2011. The CGIAR at 40 and Beyond: Impacts that Matter for the Poor
and the Planet. June 2011.
CGIAR Independent Review Panel. 2008. Bringing Together the Best of Science and the Best
of Development. Independent Review of the CGIAR System. Report to the Executive Council.
Washington, DC.
27
CGIAR Science Council. 2009. Report of the Sixth External Program and Management
Review (EPMR) of the International Crops Research Institute for the Semi-Arid Tropics
(ICRISAT), February 2009.
CGIAR Science Council. 2008a. Workshop on International Public Goods. Special Session at
AGM 08, Maputo, Mozambique, 27 November 2008.
CGIAR Science Council. 2008b. Ethical challenges for the CGIAR: Report of three studies.
CGIAR Science Council Secretariat: Rome, Italy
CGIAR Science Council. 2006. Positioning the CGIAR in the Global Research for
Development Continuum. Rome, Italy: Science Council Secretariat.
CGIAR Science Council. 2005. System Priorities for CGIAR Research 2005–2015. Science
Council Secretariat: Rome, Italy. (Reprint).
CGIAR SRF. 2011. Changing Agricultural Research in a Changing World. A Strategy and
Results Framework for the Reformed CGIAR. (Accessed 20 June, 2012).
http://cgiarweb.s3.amazonaws.com/wp-content/uploads/2011/08/CGIARSRFBrochure.pdf
CGIAR Technical Advisory Committee. 2000. A FOOD SECURE WORLD FOR ALL:
Toward a New Vision and Strategy for the CGIAR. Report presented at the International
Centers Week, 23–27 October 2000, Washington, DC.
Council on Foundations, 2011. United States International Grantmaking.
http://www.usig.org/basics/legal-requirements-private.asp
Davis, J.S., Oram, P.A., and Ryan, J.G. 1987. Assessment of agricultural research priorities:
an international perspective. ACIAR Monograph No. 4, 85 p.
De Janvry A and Kassam A. 2001. Advantages and Added Value of the Regional Approach to
Research for the International Agricultural Research System. Invited paper for the Reunion
FORAGRO Sobre Prioridades Regionales de Investigacion. Mexico: CIMMYT.
Deb UK, Bantilan MCS, Bantilan FT and Gowda CLL. 2004. Spillover impacts of sorghum
research. Pages 237-260 in Sorghum genetic enhancement: research process, dissemination
and impacts (Bantilan MCS, Deb UK, Gowda CLL, Reddy BVS, Obilana AB and Evenson
RE, eds.). Patancheru 502 324, Andhra Pradesh, India: International Crops Research Institute
for the Semi-Arid Tropics.
Deb, U.K. and Bantilan, M.C.S. 2001. Spillover impacts of agricultural research: a review of
studies. Working Paper Series no. 8. Patancheru 502 324, Andhra Pradesh, India:
International Crops Research Institute for the Semi-Arid Tropics. 32 pp.
Douthwaite B., Baker D., Weise S., Gockowski J., Manyong V. M. and Keatinge J. D. H.
(2005). Ecoregional research in Africa: Learning lessons from IITA's benchmark area
approach. Experimental agriculture, 41, pp 271-298 doi:10.1017/s0014479705002681
Ekboir, J. 2009. The CGIAR at a Crossroads: Assessing the role of international agricultural
research in poverty alleviation from an innovation systems perspective. ILAC Working Paper
9, Rome, Institutional Learning and Change Initiative.
28
Feder G, Just ER and Zilberman D. 1982. Adoption of Agricultural Innovations in
Developing Countries: A Survey. World Bank staff working papers Number 542.
Gardiner P and Chapman R. 2006. Partners in Research for Development. Science Council
Secretariat. Background paper for a workshop on Positioning the CGIAR in the Global
Research for Development Continuum, held in The Hague, 19 May 2006.
Hartmann, 2009. CGIAR reforms make research decision-making distant. Science and
Development Network. Opinion, 6 October 2009.
Harwood RR, Place F, Kassam AH and Gregersen HM. 2006. International Public Goods
through Integrated Natural Resources Management Research in CGIAR Partnerships.
Experimental Agriculture, Volume 42:375–397. Cambridge University Press.
Herdt W. Robert.2012. People, institutions, and technology: A personal view of the role of
foundations in international agricultural research and development 1960–2010. Food Policy
37 (2012) 179–190
Horton, Douglas and Mackay, Ronald. 2003. Using evaluation to enhance institutional
learning and change: recent experiences with agricultural research and development.
Agricultural Systems 78 (2003) 127–142
Kanbur R. 2001. Cross-Border Externalities, International Public Goods and their
Implications for Aid Agencies. Cornell University.
Kassam, A. H. (2006) Receptivity to social research by the CGIAR. Experimental Agriculture,
42 (2). pp. 217-228. ISSN 0014-4797
Kassam, A. H. (2003). An overview of social science research in the Consultative Group on
International Agricultural Research. The International Social Science Journal 177:441–462.
Katyal JC and Mruthyunjaya. 2003. The CGIAR at 31: An Independent Meta-Evaluation of
the Consultative Group on International Agricultural Research - CGIAR Effectiveness - A
NARS Perspective from India. The World Bank Operations Evaluation Department.
Keun Lee and Byung-Yeon Kim. 2009. Both Institutions and Policies Matter but Differently
for Different Income Groups of Countries: Determinants of Long-Run Economic Growth
Revisited, World Development, Volume 37, Issue 3 pp 533-549
Kherallah, M., Kirsten. J., 2001. The New Institutional Economics: Applications for
Agricultural Policy Research in Developing Countries. Markets and Structural Studies
Division Discussion Paper 41. IFPRI, Washington, DC.
Le Page, Lloyd. 2011. The new CGIAR structure: Perspectives on the process for CRPs and
Africa Regional integration. Presented in “Supporting a Science Agenda for African
Agriculture: A conversation among stakeholders”. June 30 - July 1, 2011, Irish Aid
Conference Center, Dublin, Ireland
Lynam John, Beintema Nienke, and Annor-Frempong, Irene. 2012. Agricultural R&D:
Investing in Africa’s Future - Analyzing Trends, Challenges, and Opportunities: Reflections
29
on the Conference. International Food Policy Research Institute and Forum For Agricultural
Research In Africa.
Maredia, M.K., Byerlee, D. (Eds.) (1999). The Global Wheat Improvement System: Prospects
for Enhancing Efficiency in the Presence of Spillovers, CIMMYT Research Report No. 5.
International Maize andWheat Improvement Center (CIMMYT), Mexico, DF.
Mausch K, Chiwaula L, Irshad A, Bantilan C, Silim S and Siambi M. 2013. Strategic
Breeding Investments for Legume Expansion: Lessons Learned from the Comparison of
Groundnut and Pigeonpea. Contributed paper prepared for presentation at the 57th AARES
Annual Conference, Sydney, New South Wales, 5th-8th February, 2013.
Monchi, Lio and Meng-Chun, Liu. 2008. Governance and agricultural productivity: A cross-
national analysis, Food Policy, Volume 33, Issue 6, pp 504-512
Nelson, Richard R. and Sampat, Bhaven N. 2001. Making sense of institutions as a factor
shaping economic performance, Journal of Economic Behavior & Organization, Volume
44, Issue 1, pp 31-54
North, Douglass C. 1989. Institutions and economic growth: An historical introduction,
World Development, Volume 17, Issue 9, pp 1319-1332
Oates, W. E. 1972. Fiscal Federalism, New York: Harcourt Brace Jovanovich.
Omamo Steven Were, Diao Xinshen, Wood Stanley, Chamberlin Jordan, You Liangzhi,
Benin Sam, Wood-Sichra Ulrike and Tatwangire Alex. 2006. Strategic Priorities for
Agricultural Development in Eastern and Central Africa. IFPRI Research Report 150.
Ozgediz, Selcuk. 2012. The CGIAR at 40: Institutional Evolution of the World’s Premier
Agricultural Research Network
Pardey, P.G. and Alston J.M.. 2010. U.S. Agricultural Research in a Global Food Security
Setting. A Report of the CSIS Task Force on Food Security. Washington D.C.: Center for
Strategic International Studies, January 2010.
Pardey, P. G., J. M. Alston, and R. R. Piggott, eds. 2006. Agricultural R&D in the developing
world: Too little, too late? Washington, DC: International Food Policy Research Institute.
Pardey, P.G., Alston, J.M., Christian, J.E., Fan, S. (1996). Hidden Harvest: U.S. Benefits
from International Research Aid, Food Policy Report. International Food Policy Research
Institute (IFPRI), Washington, DC.
Pineiro M. 2007. Agricultural technology transfer to developing countries and the public
sector. Science and Development Network, Policy Brief.
Pingali Prabhu and Kelley Tim. 2007. The Role of International Agricultural Research in
Contributing to Global Food Security and Poverty Alleviation: The case of the CGIAR, In: R.
Evenson and P. Pingali, Editor(s), Handbook of Agricultural Economics, Elsevier, 2007,
Volume 3, pp 2381-2418.
30
Pritchett, Lant and Woolcock, Michael. 2004. Solutions When the Solution is the Problem:
Arraying the Disarray in Development, World Development, Volume 32, Issue 2, pp 191-212
Raitzer, David A. and Kelley Timothy G., 2008. Benefit–cost meta-analysis of investment in
the International Agricultural Research Centers of the CGIAR, Agricultural Systems, Volume
96, Issues 1–3, March 2008, Pages 108-123, ISSN 0308-521X
Renkow Mitch and Byerlee Derek. 2010. The impacts of CGIAR research: A review of recent
evidence, Food Policy, Volume 35, Issue 5, October 2010, Pages 391-402, ISSN 0306-9192.
Ruttan, Vernon W., 1975. Technical and institutional transfer in agricultural development,
Research Policy, Volume 4, Issue 4, October 1975, pp 350-378.
Ryan J. 2006. International Public Goods and the CGIAR Niche in the R for D Continuum:
Operationalising Concepts. Revised draft of a discussion paper prepared for a brainstorming
session between some members of the Science Council of the CGIAR and the Centre Deputy
Directors’ Committee (CDDC) of the Future Harvest Centres of the CGIAR at the CGIAR
Annual General Meeting in Morocco, 3 December 2005.
Sagasti F and Timmer V. 2008. An Approach to the CGIAR as a Provider of International
Public Goods. Paper submitted as a contribution to the work of the Independent Review Panel
(IRP) of the CGIAR.
Samuelson PA. 1954. The Pure Theory of Public Expenditure. The Review of Economics and
Statistics, 36:4 (November 1954): 387-389.
Shelanski, H.A. and P.G.Klein. 1995. Empirical Research in Transaction Cost Economics: A
Review and Assessment. Journal of Law, Economics and Organization, 11 (2), 335–61.
Sharma, Devinder. 2004. Agricultural Research: CGIAR turns to outsourcing for
multinationals. In Motion Magazine May 9, 2004.
Shiferaw B, Bantilan MCS, Gupta SC and Shetty SVR. 2004. Research spillover benefits and
experiences in interregional technology transfer: An assessment and synthesis. Patancheru
502 324, Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid
Tropics. 140 pp. ISBN 92-9066-469-X. Order Code: BOE 033.
Spielman, David J. 2007. Pro-poor agricultural biotechnology: Can the international research
system deliver the goods? Food Policy 32 (2007) 189–204
Twomlow, S., Shiferaw, B., Cooper P., and Keatinge JDH. 2008. Integrating Genetics and
Natural Resource Management for Technology Targeting and Greater Impact of Agricultural
Research in the Semi-Arid Tropics. Experimental Agriculture Volume 44, pp. 235–256.
Von Braun Joachim, Fan Shenggen, Meinzen-Dick Ruth, Rosegrant Mark W and Pratt
Alejandro Nin. 2008. International agricultural research for food security, poverty reduction,
and the environment: What to expect from scaling up CGIAR investments and “Best Bet”
programs. Issue briefs 53, International Food Policy Research Institute (IFPRI).
Walker T., Maredia M., Kelley T., La Rovere R., Templeton D., Thiele G., and Douthwaite B.
2008. Strategic Guidance for Ex Post Impact Assessment of Agricultural Research. Report
31
prepared for the Standing Panel on Impact Assessment, CGIAR Science Council. Science
Council Secretariat: Rome, Italy.
Williamson, O.E. 1991. Comparative Economic Organization: The Analysis of Discrete
Structural Alternatives. Administrative Science, Quarterly, 36 (2), 269–96.
World Bank 2007. Agriculture for Development: World Development Report, 2008.
Agriculture for Development. The World Bank. Washington, DC
World Bank, 2006. Enhancing Agricultural Innovation: How to Go Beyond the Strengthening
of Research Systems. Economic and Sector Work. The World Bank, Washington, D.C.
Zeigler, Robert S and Mohanty, Samarendu. 2010. Support for international agricultural
research: current status and future challenges, New Biotechnology, Volume 27, Issue 5, 30
November 2010, pp 565-572.