Subnational Adaptation Finance Allocation: Comparing Decentralized and Devolved Political Institutions in Kenya

Adaptation finance is designed to help vulnerable populations withstand effects of climate variability and change. However, levels of vulnerability seldom determine finance distribution. Political and economic preferences of national and local government decision-makers tend to direct funding streams. This article takes an institutional approach to adaptation finance allocation by comparing decentralized and devolved local governance structures managing adaptation finance in Kenya before and after the Constitution of 2010. Prior to reforms, funding was directed through decentralized mechanisms operating within district councils and local authorities; recently, devolution of political, administrative, and fiscal decision-making to county governments coincided with piloting of new local adaptation funds. Theory suggests that devolution institutionalizes more participative decision-making and fairer allocations. Evidence suggests vulnerable communities are indeed more likely to access, design, and receive allocations of finance in devolved political systems.

The objective of adaptation finance is to assist vulnerable populations experiencing climate hazards to change behavior and support strategies to manage and lessen risk. 1 In sub-Saharan Africa, climate hazards primarily involve flooding, droughts, and storm events. 2 Yet, disadvantageous socio-economic circumstances condition exposure and sensitivity of these localities to such adverse climate events.In particular, the high poverty rates, weak infrastructure, and insufficient agency of populations experiencing physical hazards are the main factors causing high climate risk. 3Adaptation finance improves the situation by eliciting risk-reducing behavioral changes, such as livelihood diversification, water management, climate-resilient cropping, and resilience to climate disasters.
Exactly who receives adaptation finance is highly contentious.Academic and policy researchers advocate that the vulnerable participate in decision-making 1. Ciplet et al. 2013.2. Eriksen and Lind 2009; Fazlul and Nobuo 2008; Lopez-Marrero 2010; Mustafa 1998; Stringer  et al. 2009.3. Busby et al. 2013.and receive a fair share of funding relative to those with low climate risk. 4Yet, research suggests allocation decisions are made according to absorptive capacity, cost-effectiveness, or even emissions reductions 5 -factors often inversely related to climate vulnerability.
The literature has not yet considered how variation in subnational political institutions determine the distribution of adaptation finance; 6 and more specifically, how devolved funding mechanisms compare with more decentralized administrations in terms of incorporating vulnerable groups in decision-making and fund allocation outcomes.Indeed, research into the local administration of adaptation financing is now more pressing as similar initiatives funded by the Green Climate Fund (GCF) and Department for International Development's (DFID) International Climate Fund (ICF) are becoming operational across the developing world. 7his article investigates variation in local political institutions in Kenya and their function in structuring access to/allocation of adaptation finance.Do devolved or decentralized local-level political institutions facilitate greater access and distribution for the most vulnerable?
Politicians and political theorists often suggest that devolution, relative to decentralization, structures fairer outcomes through better inclusion of locallevel actors in decision-making.Decentralization involves "the transfer of public authority, resources, and personnel from the national level to subnational jurisdictions," 8 bringing administrative, political and fiscal planning, and decisionmaking closer to citizens and increasing participation.Conversely, devolution is a "statutory granting of (administrative, political, economic) powers from the central government of a sovereign state to government at subnational levels." 9evolution is a significant, authorized, and permanent transference of capacity to governing institutions with often constitutionally mandated legislative and decision-making powers.Donors and development agencies suggest decentralization ensures greater participation for citizens, improving local governance and better addressing issues for those in need. 10Devolution is a firmer, deeper, and more institutionalized transfer of political power to localities than decentralization that moves prioritization of development activities closer to those in need and structures accountability.What follows uses decentralization and devolution as analytical 11 categories by empirically investigating variation in institutional forms in Kenya, and their effects over time.
The 2010 Constitution of Kenya shifted local governance from decentralized to devolved governance, offering an opportunity to investigate institutional effects on adaptation finance.Decentralized districts and devolved county governments in Kenya have overseen local funds assisting populations with issues such as drought, flood, and land degradation.This article compares the Constituency Development Fund (CDF) 12 and Local Authority Transfer Fund (LATF), operating under decentralization, with the recently piloted Climate Adaptation Funds (CAF) associated with devolution to county government.The main finding is that vulnerable communities are more likely to access, design, and receive allocation of finance, providing a shift toward more devolved political systems. 13his article first explores the literature on adaptation finance distribution, general resource allocation and comparative environmental politics; next it establishes the comparative framing of decentralization and devolution in Kenya; then it outlines differences between devolved and decentralized governance structures.The last section compares community-level capacity to articulate demand for adaptation finance to funding structures as an indication of participation, and decisionmaking power differentials among key actors as an indicator of fair allocation.

Current Literature
Research into subnational resource distribution investigates factors determining allocation on food aid, 14 disaster response, 15 and secondary school government grants. 16Scholars tend to model allocation based on whether resources are given to administrative units of greatest need, or if elements of government interest are observable in allocation decisions.They generally conclude that resources are often given to areas with interest linkages to government, but this is far from conclusive, as indicators of need sometimes determine where money is sent.
Adaptation finance researchers primarily focus on international funding, and more specifically, that given through the Adaptation Fund of the United Nations Framework Convention on Climate Change.There is some suggestion that higher climate vulnerability 17 is associated with lower rates of Adaptation Fund distribution, 18 while others find that equity (at least between states), or allocation to the most vulnerable, informs donor decision-making. 19Equity is 12. Designed to channel funding for infrastructural and rural development initiatives to communities through local politicians.13.Pelling 2010.14.Clay et al. 1999; Jayne et al. 2001; Jayne et al. 2003.15.Besley and Burgess 2002; Francken et al. 2012; Morris and Wodon 2003; Takasaki 2011.16.Reinikka and Svensson 2004.17 generally understood by welfare economists as fairness, and has direct relations to justice concerns in adaptation finance scholarship.To date, a limited literature investigates the subnational allocation of adaptation finance, 20 and institutional factors are yet to be systematically analyzed.
This article also contributes to the literature applying comparative methods to environmental and climate change politics. 21While much of this literature compares political institutions between countries/continents, 22 it is also possible to analyze subnational institutional effects over time within a state. 23The recent transition in Kenya from decentralized to devolved subnational political institutions enables comparison of institutions structuring adaptation finance outcomes for communities while minimizing confounding factors.
What follows investigates local governance as a factor in the allocation of adaptation finance.At the same time, by investigating such institutions in Kenya, the article offers insights into broader processes of decentralization and devolution that have been ongoing across Africa for years. 24A limited scholarship traces the financing of particular programs within districts, 25 but little is known about local level structural differences in power and decision-making in relation to adaptation finance.Scholars of African politics consistently show how variation in local institutions in turn structures political and economic outcomes, 26 and the analysis below uses this variation to investigate the effects for adaptation finance allocation.
Resource Allocation in Kenya: Before and After Devolution Kenya offers a quasi-experimental setting for examining how decentralized and devolved political institutions influence resource allocation, and adaptation finance more specifically.Similar to other African states, the government of Kenya operates modern bureaucratic state institutions working in parallel with levels of customary institutions. 27Yet issues of ethno-regional patronage, absorptive capacity, climate vulnerability, and the depth of aid networks potentially determine adaptation finance allocation, 28 and such factors have also remained stable throughout the period of institutional change.
For example, the allocation literature 29 and macro-theories of African politics30 often suggest that ethno-regional patronage drives resource distribution 20.Barrett 2014.21.Harrison and Sundstrom 2007; Steinberg and VanDeveer 2012.22. Selin and VanDeveer 2012.23.Snyder 2001.24.Cameron 1995; Wunsch 1998.25.Schou and Tsoka 2010.26.Boone 2003; Mamdani 1996.27.Mamdani 1996.28.Determinants of subnational adaptation finance allocation are taken from Barrett (2014).Findings are consistent with the resource allocation literature: need, patronage, aid infrastructure, and particularly absorptive capacity.29.Clay et al. 1999; Jayne et al. 2001; Reinikka and Svensson 2004.and the provision of goods.However, since the 2010 constitutional reforms, ethnic groups in Kenya still correspond to geographic regions with little restructuring of group size or diffusion, indicating no decisive shift in inter-group alliances and patronage networks to fuse cooperation.As under decentralization, no single ethno-regional group can form a decisive majority and dominate patronage-based resource allocation since the changes.
The depth of aid networks is known to influence the likelihood of receiving adaptation finance.The scale of aid networks is a function of external assistance, and significant changes in aid levels potentially change the allocation landscape.Within Kenya, similar levels of official development assistance (ODA) are in operation within the state before and after devolution, rising steadily from $1.7bn to $2.65bn between 2009 and 2012. 31 Climate vulnerability-understood as exposure, sensitivity, and adaptive capacity to drought and floods, and driven by socio-economic factors associated with poverty and underdevelopment-is negatively related to adaptation finance allocation.In short, the more vulnerable the locality, the less likely they are to receive funding.Climate vulnerability indices 32 consistently rank Kenya as highly vulnerable with susceptibility to suffer economic and social losses-a factor that holds across the transition from decentralization to devolution. 33inally, absorptive capacity is a significant driver of adaptation finance allocation.Fund allocators require political administrations to use aid effectively, 34 which is typically understood as the ability to generate economic growth and poverty reduction. 35In the allocation literature, development levels are considered appropriate indicators of absorptive capacity, 36 or the ability to effectively use capital.From decentralization to devolution, Kenya maintained low levels of human development (from 0.472 in 2007 to 0.519 in 2012, or 145 th in the world 37 ) and low life expectancy (56.6-61.1 years from 2008-2012 38 ), and demonstrates comparable absorptive capacity across the years of focus.Therefore, the single most important factor likely to change adaptation finance allocation over this period is variation in local institutions brought about by devolution.

Decentralization to Devolution: Variation in Local Governance Institutions
Decentralization of political and economic power has a historical basis in Kenya since establishment of the republic, but consistent re-capture of decision-making authority back to central government was present from the outset. 39Majimbo, or regional government, dominated negotiations of the independence constitution (1963) between the Kenya African National Union (KANU), the Kenya African Democratic Union (KADU), and white colonial political elites. 40In particular, the desire by white elites to retain minority power resulted in the federal character of the first constitution, but the agency of politicians saw the actual implementation fall short of federalist ideals; after the first election in 1963, the KANU government began centralizing power from the regions to the center by withholding financing to regional assemblies and local authority (LA) 41 structures.

Initial Decentralization
Despite these centralizing tendencies, 42 infrastructure for local level decisionmaking and planning were forming at the outset.Decentralization began with the enactment of a policy framework 43 mandating greater involvement in investment and planning for districts and municipalities.This received further impetus in the First National Development Plan of Kenya 44 and subsequent iterations.
However, decentralized planning only began in earnest through the Special Rural Development Programme (SRDP) in 1967.District development committees formed with district development officers playing a lead role in coordination, while the early Rural Development Fund (RDF) 45 provided the finance to construct the necessary institutions for the later District Focus initiative (1983).In essence, the "district" became the unit of rural planning.Some view the Kenyatta (KANU) administration as having permitted regional and district leaders 46 to build power bases through "clientelist" networks, but much was contingent upon political support.In particular, provincial and district commissioners were simply responsible for maintaining law and order, and crucially, ensuring that districtlevel appointments completed orders from the central government. 47he establishment of the RDF provided resources for the local design and implementation of development projects, 48 but the mandate failed to develop into more substantive political decision-making functions.Further, district development officers had little control over district budgets, and most projects were either implemented by central government representatives at the district 39.Cornell and D'Arcy 2014.40.Maxon 2011.41.LA refers to "councils" in the years of decentralization-county, city, municipal, or town.42.Institute of Economic Affairs, National Council of Churches of Kenya and Diakonia 2011.43.Republic of Kenya 1965.44.Republic of Kenya 1966.45.Hopcraft 1977.46.Provincial commissioners, but also district commissioners, divisional officers, chiefs and subchiefs.47.Barkan and Chege 1989.48.Leonard 1991.

Sam Barrett • 123
Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/GLEP_a_00314by guest on 25 May 2024 level or never materialized.At the same time, LAs began to prove ineffective in what little provision of public services they oversaw.Systemic institutional pathologies emerged in the form of fund misappropriation, a resistance to public auditing, 49 and a loss of service delivery capacity.
The District Focus (1983-1988) initiative sought to finance key institutions of decentralization.First, district development officers and district development committees received budgetary powers.They began preparing "annual annex" reports for spending, as well as a "forward budget" for the medium term. 50This aligned funding with planning processes and accorded agency and status to district-level actors.Now members of parliament (MPs) had to lobby district institutions instead of central government to access funding for constituency projects.
Starting in the mid-1990s, the Kenya Local Government Reform Programme (KLGRP) 51 further reinvigorated service delivery and functionality of LAs.These had become mired in debt, overstaffed, with poor revenue raising capacity, and with no inter-governmental grant system. 52From the KLGRP came the CDF and LATF that were initially designed to address rural poverty and development in an accessible, transparent and ethical way-reducing layers of bureaucracy needed to administer funds by placing allocation powers with local level actors.

Movement to Devolution
The impetus for devolution began after successive attempts to centralize state power by the Party of National Unity (PNU). 53Recognition grew of the need to include traditionally marginalized groups, but the drive for institutional change came about primarily after the election violence of 2008.Constitutional reforms guaranteed power sharing for opposition and reduced executive power; 54 further, the Ministry of Northern Kenya and Other Arid Lands (MNKOAL) was created to develop previously neglected areas, along with other initiatives to improve social and economic rights. 55Devolution was therefore the most significant of several measures designed to institutionalize power-sharing, thus reducing the role of ethnicity in politics and improving equality for marginal or excluded groups.

Kenyan Devolved County Governments
The 2010 Constitution of Kenya provides the mandate for devolution of certain political, administrative and fiscal governmental functions to 47 county 49.Nyanjom 2011.50.Barkan and Chege 1989.51.A donor-led initiative designed to improve local government fortify local authorities in delivering services and increasing accountability.52.Devas and Grant 2003.53.Horowitz 2009.54.Kramon and Posner 2011.55.Government of Kenya 2011.governments. 56This includes significant increases in resources and budgetary control.The transition is particularly stark in Kenya due to the lack of public expenditures available to LAs under decentralization (3-5 percent). 57By comparison, under devolution county governments have received a minimum of 15 percent of treasury funding since 2013 (with medium-term development expenditure no less than 30 percent of that amount). 58evolution began in earnest in early 2013 after the election and instatement of county governors and deputies. 59Many central government functions, such as revenue distribution for public goods provision (rising to approximately 15 percent of government activities), became the remit of county governments.The precise dynamics of decision-making arrangements are not available through secondary sources, and what follows is based on interviews with county government officials in Isiolo, Kitui, and Makueni counties.
Newly devolved county governments are smaller replications of central government.Executive and legislative branches formulate policy and pass legislation.At the head of the executive branch are the county governor and deputy governor; these have appointment powers for the county executive committee, each of whom direct portfolios, including health, environment, finance, and agriculture.Beneath these personnel, chief officers implement programs as account and budget officers in charge of specific portfolios.
The executive initiates policy outputs such as the county integrated development plan, the strategic plan, and budgets and bills.New policy initiation happens through public participation forums.These are locally advertised open meetings, whereby individuals, communities, business groups, civil society organizations (CSOs), and non-governmental organizations (NGOs) articulate priorities for development plans and budgets.In theory, queries are incorporated into draft proposals and the resultant budgets.The executive then establishes a plan, budget, and implementation framework that is put to the cabinet (governor, deputies, and executive committee members), who must approve by consensus.After shaping, bills are approved and passed by legislative assembly.
The legislative assembly operates a series of committees including budget, public finance, public investment, natural resources, environment, water, and planning.Committees debate the merits of policies, and recommendations are passed on to the legislative branch, which requests amendments from the executive.Finally, the new draft is sent to the legislature for approval.
The executive sets the agenda, with scope to shape plans, budgets and bills to their preferences, though approval of county government action is the responsibility of the legislature.In addition, the legislative branch performs check and balance functions, the most prominent being composition and direction of the executive oversight committee.This committee provides the legislative branch with power to summon executive committee members to clarify issues of contention.Only in extreme circumstances does the committee have power to initiate removal of members.
As part of this new policy process, a new approach to adaptation finance allocation is being piloted through the National Drought Managementʼs County Adaptation Funds (CAFs), the most established being the Isiolo Adaptation Fund (ICAF).This operates through interactions between the County Adaptation Planning Committee (CAPC) and Ward Adaptation Planning Committees.The CAPC meets primarily to direct funding from the newly devolved CAF, 60 by reviewing proposals for suitability.
Table 1 compares the administrative hierarchy of decentralized and devolved adminstration zones.Several points of possible confusion arise surrounding past/present adminstrative boundaries.In particular, under decentralization, districts are officially the main governance unit, but in practice, multiple centers of power govern resource allocation. 61Additionally, county governments and their main administrative units-wards 62 -are at a higher scale than districts and location zones (i.e., there are fewer county governments relative to districts, and fewer wards than locations), but the issue is how they structure powers to local communities rather than the levels of aggregation/disaggregation.Further, counties (termed "county councils") were present under decentralization, and along with city, town, and municipal authorities, were collectively known as LAs.The LAs changed under devolution to include only city, town, and municipal authorities, while county councils reduced in number 60. Hesse and Pattison 2013; National Drought Management Authority, 2014.61.Demand articulation and allocation processes often involved representatives and committees from county and district administrations.62. Wards did exist in the years of decentralization, but were not a prominent administrative unit.
These were smaller political units for elected local councillors.through aggregation and were empowered with certain governmental functions, hence the renaming to "county governments." The geographical size of the administrative units needs consideration in terms of impact on access and allocation criteria.For instance, constituencies and LAs are small relative to counties-between two and twelve constituencies per county, and greater than three LAs per county.Constituencies and LAs therefore have relative ethnic, political, and economic homogenity, with corresponding reductions in zero-sum resource allocation and the overall contensiousness of resource allocation.This may influence a greater emphasis on transparency and fairness in the allocation process of institutions in county governments that mediate disparities.

Allocation of Adaptation Finance
This section explores decentralized and devolved local governance structures allocating adaptation finance.The evidence is based on over one hundred semistructured interviews from quarterly visits (2013-2014) with local government, NGOs and community discussion groups in the counties of Isiolo, Kitui, and Makueni.The comparative methodology for the funds was developed after initial scoping visits, and findings are drawn from interview responses of actors associated with local governance, but with primacy given to the experiences of community members in accessing finance.
Under decentralization, the main mechanisms for accessing interventions addressing climate variability and change were the CDF and the LATF.Often this finance was not explicitly directed towards "adaptation" or "climate resilience," but rather addressed development issues in contexts of high climate stress (consistent droughts, floods, and land degradation, among other related factors).The analysis focuses on the decision-making dynamics of the CDF and LATF financing institutions from their conception to approximately 2010. 63Since devolution, adaptation finance allocation is also being piloted through the CAF, which at the time of writing is operating in the counties of Isiolo, Kitui, Makueni, and Wajir.
The following analysis evaluates decision-making on adaptation finance allocation based on two criteria: just access and need-based allocation.In terms of just access, sound ethical procedures indicate those affected by climate impacts should be involved in decision-making processes. 64This varies from simple expression of needs of particular communities to actual representation and voting rights on decision-making bodies.In terms of needs-based allocation, the analysis centers on the degree to which local governance institutions structure 63.After this date, various changes have been made to the CDF to ensure accordance with principles of devolution-transparency, accountability, and participation (see Republic of Kenya 2013).In addition, the amounts given to the LATF are now re-directed to county governments to facilitate service delivery.64.Grasso 2010.

Sam Barrett • 127
Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/GLEP_a_00314by guest on 25 May 2024 decision-making to allocate funds for communities most in need, or those most vulnerable to adverse climate impacts.This is a moral and humanitarian perspective about the disproportionate resource distribution to disadvantage groups. 65centralized Funds for Climate Related Interventions: CDF and LATF The CDF was designed to alleviate poverty 66 by channeling funding directly to communities, and at the same time, enabling MPs to respond to demands of constituents. 67Since 2003, the CDF has been implementing infrastructural and community projects to improve rural development by using 2.5 percent of Kenya's gross annual revenue; 68 annual allocation is on average 158.3mKenyan shillings (Ksh) or $1.8m per constituency, 69 or approximately 173 Ksh ($1.97) per capita (2003-2008). 70he process begins with proposals from sub-location development committees (see Figure 1) composed of village representatives.Committees submit proposals to locational committees (for locational administration zones see Table 1), which themselves are composed of traditional leaders among other senior community representatives.If leaders approve of the terms, the submission is passed to the Constituency Development Fund Committee (CDFC) residing at the county level (or LA under decentralization).The respective MP chairs the CDFC, which also contains prominent members of the county, 71 themselves selected by the MP. 72CDFCs, via district project committees, decide the suitability of proposals to be sent for final approval by the National Constituency Fund Committee (NCFC). 73In practice, decision-making power predominantly resides with the CDFC, but the NCFC have capacity to reject proposals and withhold funds. 74In normal circumstances, the NCFC instructs the clerk of the national assembly to allocate funds via the constituency bank account.Finally, coordination of project implementation is the responsibility of the district projects committee.Overall, the CDF clearly offers villagers little opportunity to participate in decision-making.Villagers do have a means to articulate demand, 65.Chandrasekhar 1965; Rawls 1972.66. Auya and Olino 2013.67.Van Zyl 2010.68.Nyamori 2009; Auya and Olino 2013.69 but only with consent of traditional leaders and several vertically arranged institutions governing the CDF.
Precisely which proposals receive funding from the CDF is a matter of debate.Rational politicians may view the CDF as an investment in their political careers with returns spread over electoral cycles. 75This creates possible disparities in what constitutes efficiency in the CDF-politicians may aim for political returns, while it is in the interests of citizens to maximize welfare.The first part of the process resides with traditional leaders.Chiefs and assistant chiefs can deny passage to the CDFC, and there is no representation for villages in allocation decisions.Further, politicization of allocation is apparent in relations between the CDFC and traditional leaders. 76Primary and secondary evidence indicate the decisive relationship is between MPs as chairpersons of the CDFC and traditional leaders, because political linkages (e.g., voting patterns and relative influence of chiefs within district councils) tend to supersede other factors including need. 77The CDF structures provide neither regulations nor incentives for decision-makers to allocate funds to vulnerable villages or evaluate proposals based on merit.
Conversely, the LATF Act ( 1998  2 percent of national income tax with additional LA revenues, 78 which constitutes approximately 24 percent of the average LA's financial capacity towards improving service delivery and financial management, as well as addressing outstanding debt.Annual per-capita expenditure of the LATF is comparable to the CDF, with an approximate allocation of 151 Ksh ($1.72) between 2003 and 2008. 79 Community proposals for LATF funding begin with the sub-location development committee (see Figure 2).Proposals pass to select location representatives who are employees of the county council (again, known as LAs in the years of decentralization).Demand is further articulated to the LATF advisory committee, which decides on the community actions to support. 80LATF advisory committees are chaired by members of the private and public sectors, 81 but have no mandate to include village representatives in the decision-making of the committee.Regardless of the lack of community-level inclusion, the LATF has the potential to make less politicized and more technocratic and meritocratic allocation decisions, due simply to the absence of politicians in the advisory committee.
LATF distribution is a two-stage process.LAs receive the initial allocation through the ministry of finance under stipulations of the LATF Act (1998).The permanent secretary of the ministry of local government and the LATF advisory committee 82 decide on funding levels to LAs based on the following criteria: 7 percent is equally distributed among the LAs; 60 percent is given based on population size; the remaining 33 percent is allocated according to relative urban population densities. 83Therefore, the LATF at least offers transparent justification for allocation, even if no mandate exists to structure allocation to those most in need.
In theory, the release of funds within the LA depends on meeting requirements and performance indicators set by the Local Authority Service Delivery Action Plan (LASDAP), which was established in 2000-2001.The ideal is for allocation based on participatory principles and directives of the LASDP. 84he LASDAP brings together representatives from trade associations, churches, schools, and women's, handicapped, neighborhood, health and medical groups into participatory forums where citizens discuss pressing community issues.Councils organize annual LASDAP consultative meetings to facilitate the articulation of community needs and coordinate service delivery.
In practice, participation in LATF management is low relative to LASDAP objectives.Surveys of several counties 86 show few citizens being involved in project identification, and even less participate in budgeting and planning. 87ather than being an inclusive process, targeting is determined by LATF advisory committees, largely because few resources are available to develop the necessary capacity for greater community involvement and management.
Allocation criteria are further divorced from notions of equity.Community interviews suggest levels of productivity within a particular village or cluster of villages, the outcome of allocation decisions.In addition, there are negative effects of using LA as the resource transfer mechanism between government and communities.Community interviews consistently describe mismanagement of funds by LAs and capture of resources by local officials.LAs have a reputation for politicizing allocation decisions, mismanaging funds, 88 and overspending on administrative costs. 89A recent study by the National Taxpayers Association-an independent institution seeking to promote good governance in Kenya-investigated LATF allocation between 2007 and 2008 in five counties. 90They found that of the 56.7mKsh ($649k) awarded to LAs,   86.Counties: Kisumu, Taita Taveta, Nairobi, Kabarnet, Mumias, Mandera, Nyeri, Isiolo.87.Kenya Human Rights Commission and Social and Public Accountability 2010.88.Kenya Human Rights Commission and Social and Public Accountability 2010.89.Sundet et al. 2009.90.National Taxpayers Association of Kenya 2011.18.9m Ksh ($2.15m) were either inappropriately used, wasted, or unaccounted for. 91volved Fund for Climate Related Interventions: CAF Since devolution, the recently piloted CAF 92 -including the more established ICAF and other pilots in Kitui, Makueni, and Wajir 93 -propose solutions to problems associated with over-centralized management of local level public good investments, inflexible policy design and poor communication and coordination with other service providers (see Figure 3).The fund is designed to build capacity for local government and community-based actors by enabling them to "identify, prioritize and fund plans at ward, county and cross-county level for public good-type investments that promote climate resilient growth and adaptive livelihoods." 94In particular, the majority of the fund (70 percent based on figures from Isiolo County) is designated to facilitate, fund, and support plans for climate resilient development at ward level.Per-capita annual allocation in Isiolo County is comparable to the CDF and LATF-approximately 134 Ksh ($1.53) adjusted for inflation since 2008. 95n terms of access, the CAF offers an institutionalized means for communities to articulate demand compared to the CDF and LATF (see Figure 3).Ward Adaptation Planning Committees (WAPCs) involve community representatives working closely with government ministries, planners, and civil society organizations.First, transparent selection processes identify WAPC members.Every WAPC is composed of eleven voluntary ward members including two youth representatives, three women representatives, an elected member from customary institutions, one community-based organization representative, and relevant government officers, who are present but have no voting rights. 96APC procedures differentiate themselves from the sub-location committees of the CDF and LATF by requiring members to conduct participatory livelihood and local economy resilience assessments.These identify means of managing climate challenges, changing market conditions, disease, and insecurity-with particular focus on the inclusion of different groups within the ward to express challenges they face. 97WAPCs use information to develop proposals that "prioritize and design investments that promote climate resilient growth and adaptive livelihoods, based on seven criteria (see Table 2).All criteria are included on the basis that they "build climate resilience by dealing with the underlying causes of vulnerability to existing and near future climate change." 98roposals are presented directly to the CAPC, which includes representatives from the WAPCs, local governments, 99 and other stakeholder groups.This aspect further differentiates the CAF from the CDF and LATF in terms of participationdesigning structures that provide a voice to vulnerable communities-not just in the form of activities proposed, but also in terms of voting rights and decisionmaking.In terms of allocation, the CAPC authorizes all WAPC proposals on the condition they meet the first five criteria for CAF climate-resilient investments (see Table 2); 100 and WAPCs can meet the last two guidance criteria after the initial agreement to allocate funding and with technical assistance.
The CAF model aims for complete coverage within each county in order to ensure funds reach vulnerable groups.Primary evidence collected within Isiolo County suggests that, out of ten wards (seven rural and three urban), the CAF is funding climate resilience/adaptation activities in six rural wards covering 80-85 percent of the county.The only rural ward with no activities is in the central belt area, where construction of a new resort city presents problems relating to land allocation and conflict.Therefore, complete coverage of the rural areas is  the objective.Further, interviews with coordination and implementing agencies affirm that the medium goal is for full coverage in urban wards.Conversely, there is less coverage in the counties of Kitui, Makueni, and Wajir, due to their earlier stage of implementation.For instance, of the twentyfour wards, just six in Kitui so far have received CAF funds.Nevertheless, from community-level interviews in locations both in receipt and not in receipt of CAF funding, villagers consistently state that wards of low capacity or high poverty are being targeted first 101 -wards most prone to drought, poor soil fertility, and human and animal diseases.
In summation, comparisons of decentralized and devolved structures of adaptation finance allocation illuminate a positive transition to more transparent and less politized governance of local funding in Kenya.Decentralized funds have several characteristics that question their prioritization of inclusive participation and fair allocation.The CDF is heavily politicized in the character of its decision-makers and allocation criteria.Indeed, the objective of the fund was to structure MPs control over local revenues for service delivery.The LATF at least offers the potential for technocratic or meritocratic governance, and clearly a mandate did exist for greater inclusion of citizens.Conversely, LATF allocation directs funds to increase economic productivity, rather than structuring resources to those in need, and LAs were associated with fund mismanagement.
By contrast, recently devolved funding streams improve participation and structure fairer allocation outcomes.The CAF is a much simpler and more inclusive decision-making structure: one committee is constituted explicitly for ward members to assess their vulnerability and articulate demand for funds; another offers ward members representation and voting rights on county-level allocation decisions.Decision-making criteria also ensure that vulnerable groups receive adaptation finance.In short, CAF institutions structure the outcome of greater participation for vulnerable groups relative to the CDF and the LATF.Rather than making allocations based on political and economic factors of 101.This focus on areas of low capacity is also confirmed from the Makueni selection report (STARK+ 2014).

Conclusions
This article set out a comparative analysis of decentralized and devolved adaptation finance distribution mechanisms in Kenya.Elements of decentralized structures are certainly designed to facilitate participatory access and distribution based on need, such as community-level proposals, participatory forums, and localized and meritocratic decision-making.In the practice of adaptation finance allocation in Kenya, decentralized systems create multiple layers of hierarchical decision-making, increase the agency of fund allocators, and thus inhibit benefits to those most in need.Therefore, aspects of participatory access and just distribution are present, but not consistently enough to structure beneficial outcomes for vulnerable groups.New funds being piloted in Kenya since devolution-such as the CAF and ICAF-appear to have institutionalized important aspects of participatory design and needs-based allocation, and are more in accordance with the provisions and principles of the Constitution and laws on devolution.Villages and wards assess their own needs in relation to climate variability and change, have representation on key decision-making bodies that control an assigned budget and are guided by transparent decision-making criteria-all of which contribute to structuring funds towards climate-vulnerable communities instead of areas with greater economic capacity or political connections.However, the CAF and ICAF funding mechanisms are in their infancy, and time is needed to determine sustainability over the medium to long term.
What can be said with some certainty is that improvements have little to do with scale.Subnational units like districts, LAs, and constituencies are all smaller than counties in Kenya.Rather, it is the careful design of planning and coordination committees-their membership and operational rules-that structure effects in line with principles of participation and fair allocation.
Few instances provide a suitable setting to compare decentralized and devolved political institutions.Though imperfect, the Kenyan example ensures many confounding factors are held constant, which facilitates observation of the effect of institutional change.Devolution has many advocates-especially from global institutions rewarding good governance-because of the intuitive link between more mandated localized decision-making and transparent, fair and effective public service delivery.This article provides support for this intuition, at least in the area of adaptation finance allocation.
Climate finance is becoming one of the main outputs of the global climate regime.The allocation of adaptation finance to vulnerable groups that addresses climate variability and change is therefore of growing importance to policymakers, policy researchers, and academic scholars alike.This article establishes findings on an important scale of adaptation finance allocation-local governance structures-through which the majority of community-based financing passes on its way to implementation.Further work is required to empirically document best practices in adaptation finance allocation and implementation on multiple scales across the developing world.
) was established under the Kenyan central government's Local Government Reform Program to empower LAs with the direction of small development projects.The fundraising model was based on 75.Kimenyi 2005.76.See Van Zyl 2010 for a generic argument.77.Auya and Olino 2013 explain that funds address equity issues with 75 percent evenly distributed across the constituencies of Kenya, with a 25-percent allocation to the poorest.However, the focus here is on prioritization within constituencies.

Figure 1
Figure 1 Demand Articulation and Allocation of the CDF 85 78. Relevant projects include disaster risk reduction contingency and sand dams; Thomas 2011.79.Dafflon and Madies 2012.80. Interviews suggest that LA or district representatives send the proposal on to the Ministry of Water or Forest Conservation.81.Farazmand 1999.82.Including a chairman (appointed by the Minister), a fund administrating officer, the Director of Fiscal and Monetary Affairs in the Ministry of Finance (or representative) the Director of Budgetary Supplies in the Ministry of Finance, the Accountant-General of the Ministry of Finance (or representative), three persons appointed by the Permanent Secretary in the Ministry, and three persons appointed by the Minister (Republic of Kenya 2007).83.Republic of Kenya 2006.84.Thomas 2011.85.Commonwealth Local Government Forum 2009.

Figure 2
Figure 2 Demand Articulation and of the LATF

Figure 3
Figure 3 Demand Articulation and Allocation of the CAF . Comprehensive vulnerability indices include Maplecroft (2014) Climate Change Risk Atlas and Josh Busby et al.'s (2014) Climate Security Vulnerability Model 3.0.Mapping illustrates spatial interactions between social, economic, political, and physical factors, but consistent politicized disagreements over definitions inhibit objective determinations of the "most vulnerable"(Klein  and Möhner 2011).18. Stadelmann et al. 2013.19.Persson and Remling 2014.

Table 1
Comparing Hierarchies of Administration under Decentralization and Devolution *Main operating unit primarily responsible for administration budgetary allocation.
. Based on figures given by the CDF Board of national level allocation to constituencies between 2003 and 2008.70.Dafflon and Madies 2012; Republic of Kenya 2014.71.Two Councillors, one district officer, two representatives of religious organizations, two men representatives, two women representatives, one person from an active NGO, and one youth representative.72.The CDF is unusual in offering executive powers of allocation to members of the legislature.73.Republic of Kenya 2003.74.Part 2, Section 5a and 5b state that CDF disbursements shall be approved by the NCFC, and that each disbursement shall be for proposed and specified projects.
91. Makueni County Council, Town Council of Wote, Lamu County Council, Kilifi County Council, and Tana River County Council.92.The CAF model was piloted in Isiolo (2010) by the International Institute for Environment and Development using funding from UK Department for International Development and the Catholic Organization for Relief and Development Aid. 93.The Adaptation Consortium-Kenyan National Drought Management Authority, Christian Aid, the Met Office (UK), Kenya Meteorological Service and the International Institute for Environment and Development-were awarded £6.5 million by DIFD to continue and expand the operations of the CAF from 2013 to 2016.94.Hesse and Pattison 2013, 2. 95.Based on figures from the first round of costing (National Drought Management Authority 2014), but divided by the population of Isiolo rather than the "direct participants" in projects so as to maintain comparability with the CDF and LATF.96.National Drought Management Authority 2014.
97. Isiolo Adaptation Fund 2014.98.National Drought Management Authority 2014.99.The CAPC is composed of one Ward Adaptation Planning Committee representative and key government departmental representatives (departments/ministries of agriculture, livestock, environment, meteorology, and the National Drought Management Authority).100.The table is based exclusively on the work in Isiolo (National Drought Management Authority 2014).However, other pilots apply a similar "resilience to climate change" focus (STARK+ Programme 2014).

Table 2
Criteria for Guiding CAF Climate Resilient Investments the CDF and LATF, the CAF ensures that vulnerable groups receive an equal if not disproportionate distribution of finance.