LAND LAW REFORM IN KENYA: DEVOLUTION, VETO PLAYERS, AND THE LIMITS OF AN INSTITUTIONAL FIX

Much of the promise of the good governance agenda in African countries since the 1990s rested on reforms aimed at ‘ getting the institutions right ’ , sometimes by creating regulatory agencies that would be above the fray of partisan politics. Such ‘ institutional ﬁ x ’ strategies are often frustrated because the new institutions themselves are embedded in existing state structures and power relations. The article argues that implementing Kenya ’ s land law reforms in the 2012 – 2016 period illustrates this dynamic. In Kenya, democratic structures and the 2010 constitutional devolution of power to county governments created a complex institutional playing ﬁ eld, the contours of which shaped the course of reform. Diverse actors in both administrative and representative institutions of the state, at both the national and county levels, were empowered as ‘ veto players ’ whose consent and cooperation was required to realize the reform mandate. An analysis of land administration reform in eight Kenyan counties shows how veto players were able to slow or curtail the implementation of the new land laws. Theories of African politics

that focus on informal power networks and state incapacity may miss the extent to which formal state structures and the actors empowered within them can shape the course of reform, either by thwarting the reformist thrust of new laws or by trying to harness their reformist potential.
MUCH OF THE PROMISE AND POTENTIAL of the international good governance agenda in Africa and throughout the late-developing world since the 1990s has rested on rule of law reforms aimed at 'getting the institutions right'. Faith in the power of multiparty elections was thus partially offset by faith in institutions that could remove important state functions from the domain of electoral politics through legal reform or the setting up of independent regulatory commissions. 1 These 'above politics' reforms aimed at enhancing the powers of neutral bureaucracies, technocracies, and judiciaries that could restrain rulers, be they democratically elected or not. Such reform agendas attracted the fire of critics who see 'institutional fix' strategies as inherently limited. As critical legal theorists such as David Kennedy 2 and law and global governance scholars Navroz Dubash and Bronwen Morgan 3 point out, the creation of new institutions is itself a political process that takes place on the uneven playing field of existing state structure, and in any given setting, this same context will go far in shaping the practical meanings, uses, and effectiveness of new regulatory structures and laws. 4 This article argues that the fraught history of Kenya's land law reform from the adoption of the National Land Policy in 2009 to the end of 2016 provides clear evidence in support of this critique of 'institutional fixes', and of the power of an institutions-in-context explanation. We focus on the creation and working of the non-partisan National Land Commission from 2012 to 2016.
Responding to long-standing popular demands for reforms in land law and administration was a priority of the first government that came to power through electoral turnover in Kenya's multiparty era. In 2002, newly elected President Mwai Kibaki opened the door to a national debate around land policy. This culminated in a new National Land Policy in 2009, a new constitution in 2010 that contained important progressive land clauses, and the Land Acts of 2012 that were supposed to bring the constitution's land provisions to life. The centrepiece of these reforms was a National Land Commission (NLC) which was to act as an independent regulatory agency that would stand 'above partisan politics' and be autonomous from the executive and the established political elite. Most importantly, the NLC was to be a counterweight to the Ministry of Lands, which was identified by national commissions and civil society as the institutional epicentre of executive abuse-of-prerogative and land corruption. The NLC was to drive a process of land administration reforms that would redress historical land grievances, rid the land sector of corruption, and restrain the elite's voracious appetite for land and their ability to manipulate land law to their own advantage. Yet as of mid-2016, there was extensive evidence of widespread frustration arising from the blockage and even subversion of these institutional reforms. 5 By early 2016, the NLC had been reined in by a Supreme Court advisory opinion. At the end of the year, the Land Law Amendment Act of 2016 disbanded local instances of the NLC and the County Land Management Boards (CLMBs), and further curtailed the powers of the NLC itself.
Although it will take a generation or more to assess the full impact of the 2012 land laws and of the National Land Commission, frustrations and limitations of the 2009-2016 reform process have attracted great attention inside Kenya and beyond. Most observers have traced the blockages and reversals to chronic ills of Kenyan politics, including corruption, adverse incentives created by hyper-competitive party politics, and flaws in the quality and consistency of legal process and texts. 6 These critiques resonate with Africa-centred theories of neopatrimonialism or elite settlements that show how legal and institutional reform can be frustrated by patronage politics and clientelism, corruption, low institutional capacity, vested interests, inter-elite collusion, and the workings of informal institutions. Our analysis acknowledges these forces, but seeks to take fuller account of the existing structure of formal institutions in Kenya, including representative institutions, and of deep-rooted political tensions in Kenyan society around the purposes of reform. We follow political scientist George Tsebelis in placing the spotlight on the diversity and multiplicity of 'veto players' in the administrative and representative institutions of the Kenyan government who had the power to stall, check, or undercut reform, and in some instances, to use veto action to block attempts at elite capture. 7 As Tsebelis argued, and as Michael Albertus has shown recently with respect to redistributive land reforms, legislating complex legal reform in formally democratic settings with large numbers of institutional veto players is extremely difficult. The veto-player approach focuses on the formal institutional playing field (rather than its weaknesses) and highlights ways in which it empowers political and administrative actors to shape or stall the course of reform.
This article argues that effectiveness of the NLC and its county-level emanations, the CLMBs, was curbed by veto players who pushed back against either the constitutional vision of the NLC and CLMBs, or by partisan capture of the CLMBs at the local level. Potential veto players were positioned at every level of the political system from the national executive branch to the National Assembly to the new county governors and county assemblies created by the 2010 constitution and devolution. The case thus suggests that the limits of the reform vision lay in part in the hope that an independent regulatory commission, created through the domestic political process, would nonetheless be 'above politics' and thus able to somehow circumvent the multiple and myriad interests of political actors who, thanks to positions in administrative and representative institutions that were adjacent to or interlocking with the NLC, were in a position to block or veto realization of critical parts of the NLC's original mandate.
The first part of the article is a brief review of the circumstances that mobilization of civil society around this objective makes this an especially interesting case for studying a 'rule of law' reform: the kind of popular support and high visibility that can contribute to the success of such reform initiatives were present in this case, at least in pre-legislative stages of the process. The second section briefly describes the drafting and passing of Kenya's 2012 land laws, highlighting executive branch advantage at this stage of the reform process. The third section recaps the main 2012-2016 strategies of executive branch claw-back of powers that, by the constitutional mandate and the 2012 Land Acts, were supposed to go the NLC. 8 The fourth and longest section presents the bulk of the article's original empirical contribution. It identifies and analyses veto-player action to check the powers of the NLC or delay set-up of the CLMBs in eight of the 47 Kenyan countries from 2013 to 2016. The analysis is based upon fieldwork undertaken in 2016 that relied upon county-level interviews with key informants (mostly county-level administrative and political actors), as well as news reports and grey literature, to gather information on the institutional presence and functioning of the NLC and the CLMBs at the county-level. 9 Kenyan counties were selected purposively in an effort to capture some of the main lines of variation in partisan dynamics in 2013-2016 and in land politics that were expected (hypothesized) to be salient in shaping county-level politics around the NLC and the CLMBs. Based on existing research on the intersection of partisan and land politics in Kenya, we expected counties that voted for the opposition party candidates in the 2013 presidential elections to be most eager to embrace the NLC reform agenda, and thus likely to extend the cooperation and consent needed for NLC effectiveness. Conversely, we expected the NLC to encounter veto players in counties that voted for the winning Jubilee alliance in the 2013 general elections, and in counties with high levels of rural land-related conflict. The counties included in the analysis are Kiambu, 8. For more detailed accounts see Manji, Whose land is it anyway? and Bassett, 'The challenge of reforming land governance in Kenya'. 9. The research team included one MPhil Research Associate of the British Institute in Eastern Africa (BIEA) and four Kenyan post-graduate students whose degrees related to land use politics in the county-types targeted for analysis. Our structured, focused comparison of the status and workings of CLMBs in countries selected according to principles outlined above involved interviews with academics, national-level land sector actors and NGOs; secondary and gray literature reviews; and approximately three to four weeks of fieldwork in each county, focusing on interviews with county government officials, County Land Management Board members, local real estate agents and land advocacy NGOs. The generalizability of our findings is subject to limitations of geographic scope (including omission of Coastal counties), method (purposive county sampling, small number of cases in a casebased, observational study), and data sources and availability (we observed events as they were unfolding, relying on documentary sources in the public domain and purposivelyselected interviews with individuals in political and administrative office). The total number of people interviewed at the county level is approximately 60.
Bomet, Meru, Nakuru, Narok, Isiolo, Machakos, Siaya. They range from clear and steadfast supporters of the Jubilee coalition, to split counties and post-election coalition defectors, to opposition counties, 10 some of which had highly visible and long politicized land conflicts (Nakuru, Narok), while land conflicts in others were less partisan and/or more localized (Kiambu, Bomet, Meru, Isiolo, Machakos, Siaya). 11 Our findings from the eight counties reveal a greater diversity of potentially salient administrative and elected veto players, and more politically-salient land issue dimensions, than was captured in our case selection criteria and starting hypotheses. This is partly due to the devolution ushered in by the 2010 Constitution, which spurred political and partisan competition and elite turn-over at the county level. 12 It is also partly due to post-election regional tensions that strained the Jubilee coalition. Given this complexity, the county-level cases provide strong support for a veto-players focused explanation of obstacles to realization of the NLC's original mandate. Differentially motivated actors within formal administrative and representative institutions at both the national and the county level were able to stymie action by withholding cooperation or consent. To highlight the veto mechanism and its effects, the county case studies are clustered under three subheadings that highlight the institutional locus of the veto player action that appeared to be most decisive in blocking the set-up of CLMBs: the county executive, the county legislature, or the national-level administration or ruling party. In 10. Machakos and Siaya counties voted for opposition leader Raila Odinga in 2013; Bomet and Meru were Jubilee Alliance members but fell-out with Jubilee in 2015; Nakuru and Kiambu counties voted largely for the victorious Jubilee Alliance in 2013 and stuck with it through 2016; Narok was hotly contested between Jubilee and the opposition in the 2013 general election. Isiolo voted Jubilee in 2013 but was split. On regionalism and land politics in Kenya, see notes below. 11. In terms of land tenure and land issues, the study counties also represent a spectrum: Meru, Kiambu, and Machakos have much adjudicated and titled land and low levels of landrelated conflict; Siaya's land is mostly adjudicated but most landholders await transactable titles, in a setting with low levels of politicized land conflict; land politics in Narok is traditionally dominated by bitter and locally-divisive group-ranch and native/settler issues; Isiolo is mostly Trust Land with predominantly pastoral land use; and in Narok and Nakuru, settlement schemes are the focus of high-visibility, divisive land issues. An important limitation of the study is that our cases do not include a county from Kenya's Coast. On the Coast, see Karuti Kanyinga, 'Politics and struggles for access to land: "Grants from above" and "squatters" concluding, the article discusses the scope and limits of 'the institutional fix' in the case of Kenyan land law reform and of the NLC in particular, as well as possibilities opened by veto-player analysis of legislative and policy reforms in African countries.
Land law reform in Kenya: from civil society demands to legislative provision of law Land issues have been a dominant theme in Kenyan politics for the last century. 13 Today Kenya is one of the most unequal countries in Africa, with one of the highest Gini coefficients for land inequality on the continent. 14 This state of affairs is widely understood by political analysts and ordinary Kenyans alike to be the result, at least in part, of the on-going ability of powerful individuals and groups to use state power to allocate land to themselves and to politically favoured groups. Historical land injustices, land grievances, and land revendications have fuelled political mobilizations and violent conflict at key junctures in Kenyan politics since the early twentieth century. 15 These long-standing tensions and conflicts, culminating in land-related electoral violence around the 1992, 1997, and 2007 elections, all contributed to very high levels of pressure for land law reform in Kenya. 16 A series of presidential commission reports and official policy review processes clearly pointed to a chronic pattern of land abuses by the executive branch, and by the Ministry of Lands in particular. 17 During the postelectoral violence in 2008, more than 1000 people were killed and over 300,000 displaced in violence that was partly land-related. This brought Kenya to its lowest point since independence, disgracing the ruling elite and adding impetus to long-standing calls for political reform. One result was approval of a new National Land Policy in 2009, after more than a decade of civil society activism on the land issue. The main lines of the NLP were incorporated into the 2010 constitution. Kenya's 2012 land laws provided the enabling legislation to put the new principles and procedures into practice.
One of the targets of the new constitution, and to a lesser extent the 2012 land laws, was to deal with the politicized and corrupt 'den of thieves' that was the old Ministry of Lands. Although the new land laws in 2012 did less to achieve a radical overhaul of the Ministry of Lands than many had hoped (see below), some important changes were made. The Ministry was divested of some of its key land powers such as sole control over the registries, control over the allocation and management of public land, control over resettlement, and powers to revoke title deeds that were found to have been acquired illegally. 18 Many important powers of the Ministry were transferred to the NLC or to be shared with the NLC by the National Land Commission Act 2012. The NLC was to establish its presence on the ground through the CLMBs established in each county. Seven to nine members of each CLMB were to be appointed by the NLC, but were subject to approval by the new County Assemblies and the county governors. Governors were to appoint one CLMB member. 18. It was renamed the Ministry of Land, Housing, and Urban Development, and later renamed again, to become Ministry of Lands and Physical Planning (MoLPP). In this paper, we will retain the acronym MoL. 19. Kenyans have debated whether the CLMB is a decentralized instance of a technocratic arm of government, or a 'democratically decentralized' forum for making land management more participatory.
Potentially significant powers were given to the new NLC. It was supposed to manage and administer public land in the counties in collaboration with county-level land use and physical planning committees. A key responsibility was to recover public land that had been irregularly or illegally allocated, a mandate that extended to the investigation of private land that might have been acquired illegally. 20 County governments were also given unclearly defined but potentially significant powers to manage land within their counties. Most notably, they were to manage ex-Trust Land (unregistered rural land managed by the old county councils) during the interregnum between passage of the 2012 land laws and the writing and passing of a new Community Land Bill, constitutionally mandated to be completed by 2016. 21 The net effect was envisioned as a far-reaching overhaul that would bring land administration under the rule of law through the actions of the non-partisan and supposedly independent NLC, with some powers diffused through devolution to the new county governments.
National Assembly: poor drafting by design?
Ambreena Manji argues that it was soon clear that the ambiguities and limitations of the land laws would be obstacles to the kind of land reform envisioned by the land activists and civil society groups who had pushed for the 2009 National Land Policy and the 2010 constitution. The land legislation's drafting and path through Parliament in 2012 was a process that was rushed, apparently disconnected at key points from the intent of the land provisions of the 2010 constitution, and largely divorced from meaningful citizen and civil society participation. 22 This was starkly evident in the National Land Commission Bill, which was hastily drafted by a consultant who was hired by the Ministry of Lands itself. Given the centrality of the NLC to the success of the land law reform effort, the Bill was remarkably sketchy. 23  Lack of clarity about the respective roles of the old Ministry of Lands and the new NLC thus bound the NLC in ambiguous and overlapping relationships with the national executive branch, county executives, and county legislatures. Groups such as the Katiba Institute's Consortium on Land headed by the former Chairperson of the Constitution of Kenya Review Commission, Professor Yash Pal Ghai, pointed out at the time that the proposed bills would be very hard to implement. 24 Indeed, the laws were structured in ways that created opportunities for diverse actors seeking to block, capture, or prevent elite capture of the new land institutions to exert veto powers over strategic decision areas and processes.

Executive branch blockages and veto
Over the course of 2013-2016, the Ministry of Land resisted yielding its powers and responsibilities and worked to retain the mandate of the NLC within its control. The Ministry and the executive branch battled the NLC in every conceivable way, including by starving it of funds, failing to turn over relevant information, blatant obstructionism, and openly defying constitutional and legal provisions that mandated a transfer of power to the NLC. 25 Critically, between 2013 and 2016, the NLC was not able to get access to inventories of public land or land registries. This meant that it could not identify titles or allotment letters issued for holdings on public land and was thus unable to investigate the many past land allocations that were suspected to have been illegal or irregular. It was also blocked from regularizing the allocation and titling process on smallholdings, especially in settlement schemes, even though taking up this responsibility was another core objective of those who had backed land administration reform in Kenya for many years. The NLC eventually took its case to the High Court, seeking arbitration in its institutional battles with the Ministry of Land around powers of land taxation; control of the land registries, registrars, and surveyors; and control over land registration and the issuance of titles. 26  December 2015, the Court issued an Advisory Opinion that confirmed the Ministry in many of its powers, including land titling. It obliged the Ministry to 'share information' with the NLC and called for the two institutions to 'work together'. 27 A Land Laws (Amendment) Bill drafted by the Ministry of Lands in 2015 proposed to re-centre control over land information systems in the Ministry and to disband the CLMBs that the NLC had succeeded in setting up in 44 of the 47 counties. 28 It morphed into an Omnibus Bill that was described by Muhammad Swazuri, Chairman of the NLC, in his testimony to parliament on 23 June 2015, as 'undermining devolution' and 'unconstitutional'. 29 It appeared to many reformers as a complete clawback of land administration powers by the Ministry. 30 It was passed by the National Assembly and the Senate, and was signed by the president in August 2016, two weeks before the constitutionally mandated deadline for passage of this legislation.
Enactment and implementation of the 2012 land law reforms at the county level appeared to many observers between 2013 and 2016 to be a process that was slowed, and in many cases subverted, by poor planning, weak institutional capacity, and corruption. A Ministry of Devolution draft policy paper listed the main weaknesses facing devolution of land administration as weak collaboration between stakeholders, weak monitoring and evaluation, insufficient legal frameworks, inability to develop quality legislation, the challenge of attracting and retaining staff, absence of information systems, inherited staff from local authorities, unstructured public participation, uncoordinated planning, duplication and conflict of roles and functions, and tokenism in public participation. 31 A close look at county-level politics, however, supports the main argument of this article, which is that an interpretation focused on institutional structure and on institutional and partisan veto players is also possible. The new land institutions and their mandates were intertwined with representative and administrative agencies at both the national and county levels. county executive or governor, factions within the county legislatures or assemblies, and political/ partisan elites at the national level.

County executives acted as veto players in Kiambu, Machakos, and Isiolo
In Kiambu, NLC chair M. Swazuri nominated CLMB members for local vetting in 2014, but the Kiambu county executive branch refused to reveal these names to the County Assembly members for vetting and a vote. Kiambu's governor was able to block the formation of the CLMB in the face of the vigorous opposition of the majority of MCAs, who passed an unsuccessful motion in December 2015 demanding that the Governor release the names of the seven persons nominated to the Kiambu CLMB so that the process could move forward. Between 2013 and 2016, the county executive controlled land issues in Kiambu and was able to veto the involvement of other actors. One MCA, Karungo Thang'wa from Ngewa Ward, commented that '[w]hat we see is an ostensible move by the County Government trying to grab land… the names have continued to remain secret up to date'. 32 Limuru Central MCA Njenga Murugami seconded the motion to employ a legal mechanism to compel the county government to make known the names, allow the vetting and voting, and create the CLMB: 'We need the board to check what has been happening for the last three years in Kiambu County in regard to land matters and also issue title deeds'. 33 MCAs believed that the Kiambu county executive branch official refused to constitute the CLMB in order to assert unmediated control over the alienation and allocation of public land.
Public land was indeed the key land resource under government control in this county (given that there is little Trust Land; most land is registered and adjudicated). Kiambu county officials worked concertedly to identify and assert control over public lands held under expired leases, by squatters, under allotment letters, and held illegally. 34 The county executive argued that it held at least shared powers in the repossession and reallocation of public land, as well as in regulating change-of-user on private leasehold land, authorizing subdivision of leaseholds, and even the management of forest land and vacant public land. This set the stage for 32 stalemate with the NLC, which argued in defence of its mandate to administer public lands and insisted that control over land under expired leases reverts back to the central government.
Machakos County created a CLMB in 2014, but neutralized it entirely. The CLMB was constituted mostly by members of the county land administration bureaucracy who, like other land actors in local government, were pre-devolution era hold-overs. The county level lands executive and CLMB thus 'shared staff', 35 allowing the same person to act on behalf of both the county and the NLC. Furthermore, many former Machakos county councillors remained in land offices in the new county government (as was the case in several of the other counties included in this study). 36 In Machakos as in other counties, such government employees were often associated with the privatization and sell-off of public land in the transition period. After 2013, they were strategically positioned to stall or obstruct inquiry into these activities. 37 The legal mandate of the NLC made it a potential obstacle to the advancement of the ambitious development plans of the Machakos governor. As things turned out, however, the NLC did not play this role.  39 Issuing titles worked to shore up the Machakos governor's political base as well as support for the Jubilee government, but the Ministry of Land's authority to do so was vigorously (and ultimately, unsuccessfully) contested by the NLC. Mutua also teamed up with the head of the Ministry to allocate large tracks of land for development megaprojects. Over the objections of the Ministry of Agriculture, Livestock and Fisheries, Machakos county obtained 2,000 acres of public land for the Machakos City Project, aimed at attracting investors to fully serviced industrial sites on tax-free land. In another dust-up over public land, land for the Konza City megaproject (spanning Machakos and Makueni counties) was purchased by the national government from the officials of a Machakos land buying company who were later sued by members for fraudulent dealing. 40 The constitutional mandate and the 2012 land laws surely envisioned the NLC and CLMB as prominent actors in the negotiations over such land deals. However, the NLC was checked in Machakos by national and county politicians who established a tight and politically beneficial alliance around the politics of land allocation.
In Isiolo, the county executive captured and neutralized the CLMB. In 2014, the Isiolo governor and the County Assembly duly appointed a CLMB dominated by members of the governor's own sub-clan in the Borana community. This group also controlled the County Assembly. 41 The appointment was widely viewed as an example of excessive politicization of a CLMB. Although two non-Borana were added later, the Isiolo government showed no signs of willingness to constitute a CLMB that was independent of the county executive. As one local interviewee declared, 'Borana have made it clear that Isiolo belongs to them'. 42 Isiolo's county government was strongly aligned to the national-level ruling coalition, the Jubilee coalition. It campaigned on a unity agenda of overcoming the long history of ethnic and political divisions in Isiolo, bringing together the various ethnic communities with the goal of protecting community land for pastoralists. In spite of the unity pledge, after government itself did not take a visible lead on any of the major landrelated issues facing the county in 2015 and 2016, including the drafting of the Community Land Bill that would go far in determining the locus of control over Isiolo's rangelands and public lands for the next several generations. There was frustration among some civil society and political society actors and suspicion that the county elite were positioning themselves as future leading beneficiaries of the 'opening up' of Isiolo through the mega-development projects that were being planned by the central government. 43

County legislatures acted as Veto Players in Narok, Bomet, and Siaya
In Narok County, the setting-up of the CLMB was mired in disputes over its membership and was ultimately blocked by veto-actions which reflected deeper political divisions in the county. The County Assembly rejected two out of the seven members nominated by the NLC in 2014. Most of the nominees were from the majority Il Purko clan of the Maasai. Members of minority clans, including the Governor's Siria clan, which controlled the County Assembly, objected that they were not represented. Siria clan members filed a court petition in 2015, challenging the process of setting-up the board on the grounds that it lacked transparency and was not substantially representative of the people of Narok. 44 The result was a stalemate; in 2016 Narok County had a CLMB secretary who was appointed directly by the NLC in Nairobi and assisted by clerks in Narok, but the CLMB had no members. It appeared that powerful interests lodged in the county legislature sought either to ensure decisive influence over the CLMB or to neutralize it completely.
The clan divide in Narok shaped land politics as well as competition for control within the formal institutions of government. It pitted the Il Purko, who long dominated politics in Narok through their powerful position within the long-ruling Kenya African National Union (KANU) party, against several minority Maasai clans, including the Siria clan of the newly elected Narok governor. The former county council, largely controlled by Il Purko interests, had wide powers over Trust Land in the predevolution period and was deeply implicated in the land administration malpractices and misdeeds that fuelled anger against the elite. Members of the Il Purko clan were said to control as much as 85 per cent of the land in Narok. 45 Complicating matters were fractures that exist within clans and that find expression in struggles over and within group ranches. 46 These local issues shaped the evolution of the role of the NLC and the CLMB in Narok.
Narok's governor, Samuel Ole Tunai, ran for office in 2013 on the ticket of Deputy President Ruto's United Republican Party (URP) as an underdog candidate from the minority Siria clan. He was able to take the top position in county politics because the Il Purko vote was split between two rivals, and because Tunai mobilized the support of the non-Maasai, including URP-aligned Kipsigis (representing perhaps 16,000 URP votes) residing in the 1990s settlement schemes in the Mau Forest, from which Tunai himself was said to have acquired large tracts of land. Governor Tunai's economic and political fortunes were thus linked to those of the settlers.
The NLC mandate as originally envisioned allowed it to intervene in Narok's bitter group ranch disputes to negotiate compromises, revoke illgotten titles, return grabbed land, and re-demarcate boundaries. Many group ranch officials perceived such operations and functions as a threat, surely due to the very large number of complaints made against them for alleged illegalities. 47 Members of the II Purko clan, having extensive land interests in the county, were party to disputes in the most conflict-prone areas. Meanwhile, citizen complaints against influential Siria clan members were rife on Siria group ranches in the Transmara highlands.
The NLC, in cooperation with some MCAs from constituencies affected by group ranch disputes, took up some of cases in 2015 and 2016, apparently at least in part in response to local requests for conflict mediation via alternative dispute resolution. Yet in the face of local veto players, the NLC was not able to institutionalize a local presence in Narok in the form of a CLMB, and was not able to touch the most sensitive land issues. It sought compromises where possible in group ranch conflicts and avoided the most explosive land issuesthose of settlement, indigenous and minority rights, and land-grabbing in the Mau Foresteven though such matters were arguably well within its formal mandate. 48 Bomet is another case of county legislature as veto player. The Bomet Governor's strategy was evident in the first proposed slate of six nominees for the CLMB that he (rather than the NLC) submitted for approval. It included two names from the Governor's home constituency, Chepalungu, but none from the home constituency of his leading rival for the governorship, Konoin constituency. The Bomet County Assembly twice rejected names as biased in favour of the Governor's cronies, producing a stalemate around the CLMB that was never resolved. 49 The CLMB Secretary worked in an office with a couple of staffers, but was then withdrawn completely from Bomet (and transferred to Vihiga in November 2015) due to the NLC's belief that the secretary could not manage land affairs without a CLMB. 50 Bomet's land politics thus played out in the absence of the NLC, even though the leading land issues in this county fell squarely within the NLC mandate.
The particular ire by some influential Bomet County Assembly members. 52 The bad blood between the Governor and the Assembly was manifest in the several unsuccessful attempts by the latter to impeach the former. The perception of many MCAs that the governor sought to use the CLMB as a tool of the County Executive at least partially explains their attempt to veto the operation of the CLMB.
In Siaya, old-guard local elites in the county legislature played defense. The ability of both national and local level actors to use the dispensations of the 2012 land laws and the NLC to promote transparency and rule-oflaw in the Siaya land sector was checked by the old county elite of the pre-devolution era. In 2013, the county government declared its intention to register and title all land in the county within five years. It appointed a CLMB without public or political controversy in 2014, and in cooperation with the NLC, undertook an inventory and audit of all public and community land. Interviewees in county government stressed that they were 'intent on reviewing land allocations by the previous local authorities, to unearth and reverse some irregular transactions and possibly charge the responsible individuals'. 53 This would include a review of past urban and rural land allocations by the pre-devolution local authorities and inquiries into disputes over land used by the public, including disputes involving the privatization of land that had been donated to the government for local development projects in the 1980s. 54 Many of those caught in the cross hairs of the county government and CLMB's plan to review irregular land allocation were former councillors of the defunct Siaya County and Bondo County Councils. These officials were linked to the still-operating Land Control Boards, which controlled Trust Lands and the municipal land under the pre-2010 municipal and provincial governments. In Siaya as in other Kenyan counties (including Kiambu, Machakos, Narok, and Meru), several key officials under the former system now occupied strategic positions in the new, devolved county structures. 55  the head of the County Assembly's Lands Committee, all the while maintaining strong links to the LCB in his own sub-district.
According to local interviewees, land issues were the source of a split among members of the County Assembly of Siaya. The 'new broom' faction made-up of those without ties to the former council unsuccessfully opposed the bid of the standard-bearer of the old guard faction for control over the County Assembly's Lands Committee. New brooms claimed that some members of the former council and some County Assembly members were living on grabbed public plots. 56 In 2015 and 2016, the old county council elite was well positioned to veto moves by the County Assembly and the CLMB to undertake the land inventory and audit necessary to advance transparency and rule of law in the land sector.

Ruling party and the Ministry of Lands as veto players in Meru and Nakuru
The Meru county government regarded the NLC as a 'friendly institution'. 57 Meru established a CLMB in 2013 without disputes or challenges to NLC nominations. The county's own land administration bureaucracy was highly capable, technically and professionally. County officials began digitizing records inherited from the Meru Municipal Council and employed new technical staff.
Meru also provided the CLMB and the NLC with what the county government perceived as a full list of ill-gotten public lands in the county, including Meru Town, where the county government claimed that 200 prime plots belonging to the county had been grabbed by political heavyweights. 58 Meru County officials argued that the ill-gotten public land was needed for new investors, county development projects, and new county buildings. The County Executive Committee Member for Lands called upon the NLC, and in March 2016, the NLC announced that it would 'repossess [over 200 plots of] public land that have been grabbed by private developers in Meru County, regardless of the social status of the grabbers'. 59 But in Meru, it seems that having the support of county officials was not enough to ensure that the NLC could do its job.
High-profile Meru governor Peter Munya, in firm control of the Meru County Assembly, ran afoul of the Jubilee Alliance in 2015 by opposing the dissolution of Jubilee's affiliate parties, including his own Alliance Party of Kenya (APK), to form the new Jubliee Alliance Party. 60 Apparently in retaliation, President Uhuru Kenyatta and the DP William Ruto publicly backed Meru Senator Kiraitu Murungi in his bid to unseat Munya in the 2017 elections. This stand-off came to define centre-local land politics in Meru, positioning powerful national actors at loggerheads with both the Meru County Government and the NLC in this county.
The County Assembly opposed the Ministry of Lands' selective issuance of about 135,000 title deeds in Meru County as part of the Jubilee government's agenda to deliver 3 million titles countrywide before the 2017 election. In March 2016, the Ministry distributed land titles to smallholders in Imenti South, the home constituency of Munya's rival, Senator Murungi. 61 The Meru county government attempted to block this action, citing collusion between the Members of Parliament and the Ministry of Lands. 62 It was able to secure postponement of release of the title deeds twice, protesting the Ministry's actions as undermining devolution and violating the government's legal and constitutional obligation to work with the NLC. The titles were issued, however. In Meru, party politics and the Ministry kept the NLC in check-mate.
In Nakuru, the national executive and ruling party acted as veto players. Nakuru is central to national politics, and land issues are central in Nakuru. In this tightly controlled environment, a CLMB was finally appointed in May 2015, but it was largely sidelined. The CLMB in Nakuru, like other CLMBs, was underequipped and understaffed and was not visible in land matters. One notable exception occurred in June 2016 when the Nakuru CLMB secretary himself was suspended and alleged to have been involved in corrupt land deals.
Some Nakuru MCAs appealed to the NLC in 2014 for support to address long-standing, festering issues around land holdings in former land-buying company areas in Kuresoi (Chepakundi, in Olenguruone Division), where titleholders who were evicted in 1992 election-related 'ethnic clashes' wanted to be properly bought-out by new occupants. 63 Yet as was the case for other highly sensitive Nakuru land cases involving violent land displacement and restitution after electoral violence, the NLC and the CLMB remained were on the sidelines as national-level political actors took the lead. A Parliamentary Commission was appointed in September 2015 to look into the Kuresoi disputes, for example. This is not the outcome expected in 2012, when important powers over the management of public land, settlement and resettlement, and IDP issues (and land titling) were to be shared between the Ministry of Lands and the NLC, or transferred to the NLC. Yet in line with the national government's desire to maintain control over a potential hotbed of violence, national-level actors linked to the executive and ruling party took the initiative in managing land issues in Nakuru, blocking the NLC in most domains. NLC chairman Swazuri was, however, brought in to lead 'alternative dispute resolution' efforts involving Maasai claims to lands wanted by politically well-connected project developers around Naivasha and along the Nakuru-Narok border.
The long history of government allocations and re-allocations of land meant that the NLC mandate to review all public land allocations threatened many established interests at both the county and the national level. Interviewees in county offices in Nakuru believed that most public land in Nakuru town (estimated population of 393,000 by 2017) had been grabbed by former county councillors, many of whom occupied strategic posts in the new county government.

Conclusion
The legislative effort to enact the reformist land-related provisions in the 2010 Kenyan constitution was focused on vesting land-administration authority in an independent regulatory agency, the NLC, and on leveraging democratic decentralization by giving county level actors the power to approve (veto) NLC-appointed members of the CLMBs. Yet as Dubash and Morgan and others have argued, strategies that hinge on the operations of non-partisan regulatory agencies often underestimate the importance of institutional and political context in shaping the actual outcomes. Inevitably, they argue, the drafting and implementation of reformist legislation will be shaped by previous patterns of state-building, the power of existing elites, and the interplay of interests around both the procedures and substance of reform. This has been the case in Kenya. An analysis focused on the overriding influence of informal institutions, corruption, neopatrimonialism, and/or low state capacity does not do justice to all the mechanisms at work in this case.
Power to implement land administration powers defined under the 2012 land laws lay in the hands of diverse institutional actors in both administration and representative institutions at both the national and county levels. Some of the individuals expected to implement the new land laws and cooperate with the non-partisan NLC were precisely those targeted by reform, and they often used veto powers to fight back. Others sought to expand county land prerogatives at the expense of the centre, as seen in both Kiambu and Meru, or to assert a regional political agenda at the expense of partisan players in the national ruling coalition (Bomet). Some sought to use local veto powers to block local elite capture of the CLMBs. These actors were able to use the prerogatives and powers of institutions within which the NLC and the CLMBs were nested, or of interlocking institutions at the national or country level, to neutralize, stall, or veto the functioning of the NLC and/or the creation of the CLMBs.
David Kennedy has underscored the limits of rule of law to achieve overtly political redistributive choices. Manuel Tedoro and Anne Pitcher have made the same argument about regulatory agency reform. 64 Redistribution is indeed a large part of what has been at stake in land law and land administration reform in Kenya. Constitutional provisions and new laws were to transfer power from the executive and to place it in what were envisioned to be politically neutral, more technocratic institutions, the NLC and the CLMBs, whose local credibility was to be sanctioned by elected county governments. The new land institutions were to have been empowered to recover ill-gotten public lands and stolen community lands and, in the broadest reading of the original mandate, redistribute land to those with legitimate claims and/or those wronged in the past. The high stakes of land law reform Kenya and its redistributive potential seem to make the cautionary admonitions of Kennedy and of Tedoro and Pitcher especially appropriate in this case. As Albertus showed in his study of redistributive land reforms in Latin American countries, redistributive land reforms Latin American countries, veto players can use formal state institutions including representative institutions under democratic regimes to deflect reform drives. 65 Institutional players may use their veto to block attempts to hijack reform, or to block their rivals' ability to gain advantage from reform.
The case of Kenya's 2012 land law process and the struggle over the NLC mandate underscores the persistent and perhaps growing need to take account of systemic political forces such as partisan rivalry, clashing preferences over state structure, and bottom-up pressures for redistribution in the analysis of constitutional, policy, and legislative change in African countries. While clientelism, informal political relationships, and particularistic elite interests surely play a strong role in explaining outcomes, an institutional veto player analysis such as the one developed in this study directs attention to ways in which these systemic political forces find expression in politics that are played out through formal state institutions, including representative institutions. In Kenya and many other African countries, the salience and complexity of such formal institutions has increased over time with multi-partyism, decentralization and devolution, and more open political arenas.
Veto player analysis does not propose a general theory of actors' relative power or motives, however. Here the analysis of the 2012 land law reform process in Kenya seems to underscore above all the loss of initiative by the grassroots forces rallied around the 2010 Constitution's promise of land administration policies that would recover stolen public lands and right past abuses of executive power. Kenyan reformers and their international allies threw their post-2009 efforts into a legal reform process, rather than investing equivalent effort and resources in building grassroots political momentum or political-party agendas around the 2012 legislative effort, or around support for elected county-level representatives who would tackle the substantive issues that animated land politics in particular counties. Although urban-based professional civil society organizations consulted widely with citizens at the grassroots in the run-up to the drafting of the National Land Policy, after 2009 there was no institutionalized mobilization from below around substantive land demands. We can speculate that the fall-off of popular mobilization at the legislative stage, coupled with the lack of an institutionalized movement such as a broadbased political party committed to the land law reform agenda, may have helped to empower the institutional veto players at the national and the country level who have worked to stall or thwart the constitutional promises of land law reform.