A critique of the newly proposed model of decentralisation in Lesotho

The Government of Lesotho has drafted the Local Government Bill 2016. The Bill is supposed to be a reformist piece of legislation to transform not only the model of decentralisation established by the Local Government Act of 1997, but to also undo the colonial legacy of de-concentration. To this end, the government adopted a new decentralisation policy in 2014, which unequivocally provides that the model of decentralisation in Lesotho shall be ‘devolution’. Hence, the 2016 Bill is assessed in this light: whether it implements this policy decision. This paper contends that, while an attempt is made by the Bill to introduce some semblance of autonomy for local authorities, the model of devolution it introduces is in fact very weak and needs significant strengthening if decentralisation is to succeed.

'Nyane Decentralisation in Lesotho CJLG 2019 3 deconcentrated central government departments that were still present alongside local authorities. These administrators typically have wide-ranging powers.
According to Olowu (2003, p.48) "control mechanisms include inspectorates, approval processes for local government decisions, deployment of central level personnel to local governments and use of powers of suspension and dissolution of local government councils". Some scholars have characterised this as a 'dual model' of decentralisation (Mawhood 1993). According to Mutizwa-Mangiza (1990, p. 429), the model is made up of "[on the one hand] central government's field administration, itself subdivided into quite separate ministry hierarchies operating largely independently of each other. On the other hand … local authorities which operated without any reference to central government's field administration". In that way, local governments were used pre-eminently as tools of administrative management (Ribot 2002). This colonial legacy endured long after decolonisation (Therkildsen 1993).
One of the characteristic features of this approach to decentralisation, amongst others, was that decentralisation had very weak constitutional and legal framework (Olowu 2012;Fombad 2018). It is only recently that many countries across Africa have entrenched decentralisation within their constitutions with explicit recognition and protection for local government, including Ethiopia, Ghana, Mali, Namibia, Nigeria, Senegal, South Africa and Uganda among others (Ribot 2002). There is therefore a discernible wave throughout Africa to deepen decentralisation as part of the broader constitutional and political changes sweeping across the continent. The essence of this trend is not only to give more power to the local structures and populations but also to protect them against interference from central governments.
The underlying justification for decentralisation in Africa is that it enhances democratic governance and service provision (Oyugi 2000). As Smoke (1999) pointedly contends: " [u]nderlying most of the purported benefits of decentralization is the existence of democratic mechanisms that allow local governments to discern the needs and preferences of their constituents, as well as provide a way for these constituents to hold local governments accountable to them" (p.10). With the dire state of development in which the majority of African countries found themselves after independence, it is argued that the most effective means for delivery of services and to ensure real participation of the people in public processes is through decentralisation. This presupposition has been firmly held even by multinational institutions supporting decentralisation such as the World Bank and the International Monetary Fund, "as part of downsizing central governments" (Ribot 2002, p. 1). Consequently, sometime African countries have been under pressure to decentralise as a soft condition for aid and technical support in the broader governance reforms (Smoke 1993). In equal measure, the current decentralisation reforms in Lesotho are taking place against this broader Africa context.

Constitutional basis for decentralisation
When Lesotho returned to electoral democracy in 1993, a new constitution was adopted. Broadly, the constitution recreated the country as a unitary state. While power was separated into executive, judicial and legislative branches, the constitution never divided it between tiers of government; nor did it allocate any functions to sub-national levels. Indeed, sub-national structures were only nominally envisaged, as Section 106(1) of the constitution indicates: Parliament shall establish such local authorities as it deems necessary to enable urban and rural communities to determine their affairs and to develop themselves. Such authorities shall perform such functions as may be conferred by an Act of Parliament.
The constitution thus delegates the power of decentralisation to central governmentnotwithstanding a settled view in academic circles that countries decentralise best when it is the constitution, not central government, which subdivides tiers of government and allocates general functions to each tier.
Arguably, one of the reasons why Lesotho's decentralisation has never really gained traction since 1993 is because of its weak constitutional basis. According to Fombad (2018), strong constitutionalisation of decentralisation has three main advantages. The first is that decentralisation becomes entrenched in the design of the country. It cannot be changed on the whim of those in power. Secondly, when the constitution creates tiers of government and allocates functions to each tier, it creates clarity over responsibility and authority, which protects the sub-national tiers of government. Often, otherwiseas is the case in Lesothocentral government runs roughshod over the sub-national spheres. Thirdly, constitutionalisation of decentralisation makes implementation mandatory. It does not rely on the discretion of central government. As such, "[c]onstitutional entrenchment can have the effect of legally compelling the executive and legislature to fulfil their duty to implement the system of decentralization in the exact manner contemplated by the constitution" (Fombad 2018, p. 196). Thus, a genuine conversation about the reforms of decentralisation in Lesotho may have to start at constitutional level.

The current structure of local government
After a long-drawn-out conversation about decentralisation, starting in 1993, the government enacted the Local Government Act 1997. Interestingly, the Act uses the term 'local government' as its titlebut this term is never used by the constitution. The constitution instead uses the phrase 'local authorities'. A false impression was thus created by the Act that Lesotho, like many other countries, has both a central and sub-national 'governments'. While the two conceptslocal governments and local authoritiesare sometimes treated interchangeably in the literature (Callanan 2003, p. 3;Roche 1982), they can imply different things. The main difference can be in their powers and levels of decentralisation. 'Local government' implies stronger devolution of powers than 'local authority'.  Olowu (1988, p. 12) also contends that local government must be further distinguished from other subnational structures: Local governments are identified by certain defining characteristics. These characteristics usually focus on the following five attributes: legal personality, specified powers to perform a range of functions, substantial budgetary and staffing autonomy subject to limited central control, effective citizen participation, and localness. These are regarded as essential to distinguish it from all other forms of local institutions and also ensure its organizational effectiveness.
Local authorities, on the other hand, may come in several forms; they may possess some of the characteristics of local governments but they may lack substantial political, budgetary or administrative autonomy. They can largely controlled by central government. These types of local authorities are therefore normally a type of decentralisation favoured by unitary states, such as Lesotho.
However, like the Local Government Act 1997, the 2016 Bill uses the two terms interchangeably. The 1997 law created a three-tier local structure: community councils, district councils and urban/municipal councils. Initially there were 128 community councils, 10 district councils and one municipal council (Mofuoa 2005;Daemane 2011), but the number of the community councils has been substantially reduced by half to 64 (see Table 1 below). This reduction is due to lack of resources and to improve efficiency, but has led to councils being criticised for not being local enough (Kapa 2009). Interestingly, although the local government structure during the military junta was criticised for its lack of election componentit used the amorphous process of pitso to select people into the structures 1it has nevertheless been credited with one main advantage over the current structure: that it was 'truly local'.
The structure known as the village development council (VDC) during the military dispensation, unlike its modern-day counterpart the community council (CC), was smaller in size and closer to the population it intended to serve. As such, Kapa (2009) contends that: because every village in Lesotho had its own VDC, hence offering scope for the people to have a better and direct input in issues affecting their lives, the Lesotho local government structure under the military regime was a better popular participation maximization model than the current one. Additionally, the model was a better means for service delivery since the villagers accessed services in their own villages (p. 9).
The modern-day community council, by contrast, is too big to relate directly to the needs of the concerned locality. Although electoral divisions do exist, these are not local authorities; they are only used to elect a representative of the locality onto the bigger community council.

The current political and administrative decentralisation
It is instructive to assess the extent to which political and administrative powers have actually been In that way, there is some semblance of political decentralisation. However, the local authorities do not have any meaningful executive or legislative powers, nor do their chairpersons have any political powers. Instead, they are executively and politically beholden to the Ministry of Local Government which, by and large, determines their executive programme.
The degree of a country's political decentralisation can be measured by two primary indicators (Kovar 2008). The first is the extent to which local governments are able to make decisions and set policies and rules in their areas of competence. Under the 1997 Act, local authorities in Lesotho do not have full powers to make policies and rules. The law empowers them (under Section 44) to make rules and policies, but subject to the approval of the Minister of Local Government. As a result, local authorities do not currently have by-laws because the drafts they once produced are permanently awaiting the approval of the minister. The second key indicator of political decentralisation is the level at which power is transferred away from central government. In Lesotho, the powers of councils are transferred by statute (lower level), not granted by the constitution (higher level). Furthermore, the Minister of Local Government has excessive powers under the Act, including the power to dissolve a council (Section 25). This model of administrative decentralisation runs is the weakest model because the local authorities have powers on administrative staff (Schneider 2003). Since in Lesotho executive officers of councils are neither recruited nor controlled by the councils, this suggests that de-concentration is being perpetuated.
Nor is this all. Several central government ministries (Education, Finance, Agriculture and many others) also maintain a presence at local level that is completely uncoordinated with the local government structures. In a bizarre and unworkable attempt to square this circle, the Local Government Act 1997 establishes the office of district administrator as "the person who shall represent the interests of the central government at district level and shall be responsible for coordinating the duties and functions of all public officers in that district, other than those employed by local authorities" (Section 39 (2)).
The retention of a foothold by central government at local level is a phenomenon common to most former British colonies in Africa (Oviasuyi and Lawrence 2010). In Lesotho the legacy of duality was actually strengthened in the aftermath of independence. Due to the avid belief in recentralisation at the time, the government pushed most central departments to have a presence in the districts with a view to plugging the gap left by the disbanding of district councils which had taken place (Mofuoa 2005;Daemane 2011). The position of district administrator, which has been so enduring in contemporary local governance, was consolidated during this era to represent the central government at district level (Kotze 1972). In fact, district administrators were directly responsible to the Office of the Prime Minister.
It would seem, therefore, that right from the conception of the current design for local authorities in Lesotho, the central government always planned to maintain a permanent deconcentrated presence at local level, despite its original Concept Paper envisaging a local government system "based on political devolution and decentralisation of functions, staff and finance" (Government of Lesotho 2004, p. 5).
The current design is such that despite the Councils being headed by chairpersons as political heads, the chief executive officers, who are secretaries, are not accountable to the political heads of the councils (see Table 2 below).

Fiscal decentralisation
Fiscal decentralisation is the bedrock of any decentralisation project. It can be defined as the assignment of revenue and expenditure functions to sub-national tiers of government (de Mello and Luiz 2000); and the extent of fiscal decentralisation shows the extent to which a country is prepared to devolve practical policy-making powers to the local level. There are three key measurement tests to determine degrees of fiscal decentralisation (Kovar 2008). The first is the extent to which local authorities have sufficient resources to execute functions assigned to them. The second is the extent to which government funding for local authorities is stable and predictable; and the third is the extent to which local authorities are able to generate own revenues at local level. In Lesotho, fiscal decentralisation is the weakest link in the decentralisation policy. While the 1997 Act empowers local authorities to collect local revenue in the form of tax and rates (Sections 47 and 56), the Minister is simultaneously mandated to specify the list of items over which councils may collect revenues by way of "tax or levy or service charge" (Section 57). Over 20 years later, there is still no such list, so consequently local authorities are not collecting any rates or tax revenue.

Decentralisation policy 2014
The f) implement a strategy for human resource management in local governments; and g) provide an enabling legal framework for devolution.
As such, it is clear what kind of decentralisation the policy seeks to attain: a shift from a local government beholden to central government to one that is autonomous. This shift can best be realised through legislation intended to give local authorities political, administrative and fiscal autonomy.
Therefore, this is the prism that must be used to assess the Local Government Bill 2016: does the Bill achieve the aforementioned policy intentions?

Type of decentralisation prescribed by the policy
The 2014 policy categorically states that the model of decentralisation envisaged is 'devolution'. It even defines what devolution isdespite this being a fairly settled concept in local governance studies (Conyers 1983)namely, as: the process of transferring functions, responsibilities and resources from the central government to lower tiers of government by legislative provisions so that the entities are legally empowered to make decisions on their own. The transfer is within formal political structures and is institutionalized by constitutional means... Under this form of decentralization leadership is accountable to the local population through a system of elections (Government of Lesotho 2014, p. 37).
The policy calls strongly for a constitutional and legislative framework for devolution. However, devolution can still vary according in degree, on a continuum from strong to weak. differences between strong and weak devolution. Comparing the policy to the factors in Table 3 evinces the argument that it envisages a strong rather than a weak model of devolution. Source: Kovar (2008)

Structure and intergovernmental relations model proposed by the policy
The policy proposes to change the current structure of local government and the relationship between the central and local governments. In contrast to the current structure, where the political head of a local authority is a non-executive chairperson, the policy proposes that the political head of a local authority should be a mayor with executive powers, and the chief executive officer accountable to the mayor would have the title of district executive secretary (DES) (see Table 4). The mayor would be responsible to the local council, which would be created as a 'local parliament' with capacity to make by-laws and provide checks and balances through political oversight of the operations headed by the mayor. In this way a local authority is envisaged as a true government with powers separated between the executive (headed by the mayor) and the legislature (headed by a chairperson or speaker). The policy also stipulates that the council should comprise directly elected councillors and persons specially nominated on the basis of their expertise, experience and specific interest in the development of the area. This duality is intended to combine expertise with local democracy.
While the policy provides that the political head of a local authority should be a mayor, the influence of central government is still evident because, contrary to its definition of devolution, the policy still  Table 3 above, when there is too much interference by central government in local affairs, devolution is weak.

The structure of local government proposed by the Bill
The Bill was drafted in 2016 with a view to implementing the 2014 decentralisation policy. As outlined above, the policy unequivocally envisages decentralisation based on devolutiona radical departure from the current model, which is based on a vague mixture of devolution and deconcentration. However, this radicalism is not followed through in the Bill. Instead, it provides for five tiers of local government, namely: a) municipal councils; b) city councils; c) district councils; b) urban councils; and e) community councils (Clause 3). But only one of thesecity councilsis new. Under the existing structure, a major criticism has been that community councils are too big to respond adequately to local needs. The electoral division is a more 'local' grouping, but electoral divisions are not local authorities. They are merely used to organise elections to community councils. Another key weakness of the Bill is that it does not define the characteristics of councils, but rather delegates their definition and establishment to the Minister of Local Government. This not only undermines the notion of devolution, but also borders on violation of Section 106 of the constitution, which provides that "[p]arliament shall establish such local authorities as it deems necessary to enable urban and rural communities to determine their affairs and to develop themselves" (Section 106(1)). As the Bill specifies the names of the different council typesalthough without detailing the characteristic features of the various categoriesit usurps the But, as demonstrated above, according to the constitution the power to 'establish' a local authority belongs to parliament.

Political decentralisation model proposed by the Bill
As discussed above, political decentralisation is the bedrock of any decentralisation process. It indicates the extent to which a central government is ready to allow sub-national structures to make political decisions relating to policy, planning, human resources and finances. Furthermore, the sub-national structures, to be fully autonomous, must be elected by local people, not appointed by central government (Fombad 2018). On this point, the Bill is definitely an improvement on the existing situation, as it creates councils with separate executive and legislative functions. It proposes that the executive authority of the council will be vested in an executive committee (Clause 38), led by an executive mayor who is elected by the council. This is a welcome development, and contrasts with the existing set-up in which the political heads of councils are non-executive chairpersons, and councils lack clear executive responsibilities. Indeed, at an administrative level, councils are currently headed by a secretary who is not controlled or even appointed by the council. Furthermore, at the legislative level the Bill seeks to establish 'local assemblies'. These assemblies are to be responsible for making by-laws and overseeing the work of the executive committee. This is also a welcome development because, although the existing legislation gives local authorities power to make by-laws, these are subject to approval by the minister. The Bill (Clause 54) replaces 'approval' with 'certification':

(1) A by-law, amendment, variation or rescission of any by-law under this Act shall have no effect until -(a) the Minister has certified that it does not conflict with any national law; and (b) the certification is published in the Gazette. (2) A by-law made under this Act shall be published in the Gazette with the certification required under subsection (1) (a) and it shall take effect upon such publication.
It remains to be seen, however, whether in practice the process of 'certification' will differ from 'approval'.

Administrative decentralisation model proposed by the Bill
Administrative decentralisation concerns the extent to which a central government is willing to devolve delivery of public services to local structures. It relates, in the main, to the appointment and functions of non-political human resources. As demonstrated in Table 3  Bill. The implication is that councils will be empowered with political and executive functionsbut cannot be trusted with employment and control of their own staff.

Fiscal decentralisation model proposed by the Bill
Fiscal decentralisation consists primarily of devolving revenue sources and expenditure functions to The advantage of the model was its objectivity; it removed a lot of discretion from the minister. This was important because the main problem with fiscal decentralisation in Lesotho is that there is a widespread reluctance on the part of central government to release financial resources to local authorities. Consequently, the Ministry of Local Government ends up being a de facto guardian of local authorities. Unfortunately, the new Bill seems to perpetuate, rather than end, this practice. Clause 63 provides that: (1) A Council may, in order to implement its development programme, make a written application to the Minister for the allocation of a grant.

CJLG 2019 14
(2) An application under sub-section (1) shall be submitted in such form as may be prescribed by the Minister, containing such particulars, together with such documents as the Minister may determine.
(3) The Minister shall evaluate the merits of such application having regard to the development priorities of that Council, and recommend such application to the Minister responsible for finance.
(4) Where the Minister of Finance considers such application and allocates a sum of monies by way of grant in respect of an application made under sub-section (1), such monies shall be credited to the Fund of the Council.
Clearly, therefore, the Bill does not introduce any statutory intergovernmental transfer model. Instead the minister, as is the pattern throughout the Bill, is left with wide discretion to determine the transfers to local authorities. This is most unsatisfactory and will not contribute to strong decentralisation.
Intergovernmental transfers are generally the main source of funding for local authorities. International studies of local governance patterns indicate that only city councils or metropolitan councils are able to fund local development through locally generated revenues, while local authorities that are not city municipalities or metropolitan councils rely on intergovernmental transfers for almost 90% of their revenue (Bahl 2001;Dziobek et al. 2011). With this in mind, it seems inconceivable that new legislation on local governance in Lesotho should not seek to regulate this vital aspect of fiscal relations between central and local governments. What is needed is an objectively verifiable formulaor at the very least, mandatory guidelinesto deliver intergovernmental transfers.

Functional assignment model proposed by the Bill
The extent to which a country is prepared to assign functions to local authorities is another key indicator of its commitment to decentralisation. As demonstrated earlier, in many countries this assignment is done at constitutional level. In Lesotho, however, assignment is delegated to parliament (Section 106 of the constitution of the Local Government Act 1997): Every local authority shall, subject to the powers reserved to or vested in any other authority by this Act or by any other written law, be the authority within its administrative limits charged with the regulation, control and administration of all matters as set out in the First Schedule (s 5(1).
The Local Government Act allocates functions to local authorities via schedules to the Act (see Table 5 below). For example, under Section 5(2), community councils perform exclusively the functions in the Second Schedule. The schedules have been criticised for their generality ('Nyane 2016), but at least they attempt to demarcate functions for local authorities. An improvement to this model of decentralisation would be to provide more detail on the rather generic categories below. For instance, it is provided that all local authorities have competence to handle 'education', but it is not clear whether education includes 'higher education'. The same lack of detail also affects other functional areas, such as 'land'.  However, one crucial change is that the new Bill completely removes responsibility for functional assignment from parliament, and instead hands it to the minister. Clause 7 of the Bill provides that: (1) The Minister shall - (2) A Council shall, after the transfer of responsibilities and resources contemplated in subsection (1), be the authority within its jurisdiction charged with the regulation, control and administration of all matters related to the relevant functions as defined in terms of subsection (1).
This formulation has two major problems. The first is that it violates the constitutionwhich not only empowers parliament to establish local authorities, but also provides that "[s]uch authorities shall perform such functions as may be conferred by an Act of Parliament" (Section 106 (1) The second problem with the model of assignment under the new Bill is that by empowering the minister, rather than parliament, to decide functional assignment, the aim of Lesotho's decentralisation policy is undermined. The policy is clear that the model of decentralisation to be followed in Lesotho is 'devolution'and, since local governance is weak at the constitutional level, a bold Act of Parliament is needed to devolve powers to local authorities. By contrast, allowing the minister to assign functions instead strongly suggests the government is favouring another model of decentralisation (ie deconcentration)which is not the one envisaged by the policy (Government of Lesotho 2014).

The role of chieftainship under the Bill
The role played by chieftainship is another unresolved question of local governance in Lesotho. Despite consistent weakening of chieftainship since colonial times, the institution seems to be resilient and remains tenacious in local governance of the country (Leduka 2006

Conclusions
The drafting of the new Bill on local governance in Lesotho comes after several decades of 'trial-anderror' decentralisation in the country ('Nyane and Tsikoane 2017), from colonialism to the current dispensation anchored by the 1993 constitution and the Local Government Act 1997. The government has acknowledged that the present model of decentralisation is based on deconcentration (Government of Lesotho 2014) or at least a tacit combination of deconcentration and devolution, and concedes that the project in this form has not worked; and there is therefore a settled intention to move to a new approach based on devolution and increased autonomy of sub-national governments. However, several aspects of the Bill seem to entrench a very weak type of devolution. Firstly, as discussed above the Bill seems to violate the constitution by empowering the minister to 'establish' local councils and 'assign functions' to those councilspowers which the constitution has explicitly mandated to parliament. This is tantamount to usurpation of parliamentary powers. Secondly, the administrative decentralisation proposed by the Bill is weakened by the fact that senior administrative personnel in councils are to be appointed by the Minister of Local Government, not the local authorities where they will work. This approach seriously undermines the autonomy of local authorities. Thirdly, the model of functional assignment proposed is a backward step which will only perpetuate the present deconcentration model. At least the present legislation sets down schedules of functions for local authorities; but these would be removed under the new Bill, with powers shifted to the minister instead.
This will mean local authorities will depend on central government for both their existence and functionswhich as has been demonstrated above results in the weakest possible model of decentralisation. Fourthly, fiscal decentralisation is particularly weak under the Bill. Instead of improving on the present model, the Bill retains it, such that local authorities remain beholden to the minister for grants and transfers. It is widely accepted that intergovernmental transfers constitute the most important, if not the only significant, stream of funding for local authorities. Therefore, this paper argues that these transfers must be regulated by a clear formula so that local authorities are not at the mercy of central government.
Finally, the Bill fails to introduce needed changes to the system of chieftainship. Its approach only serves to momentarily pacify the institution; it pays lip-service only to the institution of chieftainship as part of local governance. This is most unsatisfactory, as it will not address the violent tensions that currently exist between elected and hereditary levels of local authority.

Recommendations
In light of the foregoing analysis, the following recommendations may be made: Constitutionality of the Bill: There is a real need to review Section 106 of the constitution in order to enhance decentralisation at constitutional level. However, the immediate priority should be to ensure the Bill does not contradict existing constitutional provisionsie that it specifically establishes local authorities, their categories and their functions, rather than leaving these at the discretion of the minister.
Administrative decentralisation: The power to recruit and control senior officers of local authorities must be removed from the minister. The Bill does increase the executive and political powers of local authorities, so it is theyin collaboration with a properly structured Local Government Service Commissionwho should be responsible for recruitment and control of their professional staff. Functional assignment: Decentralisation works best when the functions of sub-national structures are enshrined in law, rather than 'gifted' by central government. The Bill must be reviewed to include schedules of functions for various categories of local authorities, and the opportunity should be taken to make them more specific and clearer than the present schedules under the Local Government Act 1997, in order to properly guide local authorities in their jurisdictional competences.
The role of chieftainship: Lesotho has yet to start a meaningful conversation about the substantive (as opposed to formal) role that chieftainship should play in the decentralisation project. However, in the meantime this paper recommends that the Bill includes a schedule of functions for chieftainship, which is drawn up following consideration of both the existing customary functions of chiefs, and those mandated under the Chieftainship Act 1968.