Is Election a Disadvantage ? Nigerian Local Councils and Security Provision

Deepening democracy through local elections is a major aim of decentralisation. But do elections always work to strengthen the relationship between local politicians and citizens, or can they set elected officials up for failure? This article examines security provisioning by local governments in Nigeria, understanding that some local governments are elected, while others are appointed by state governors. Our hypothesis was that elected local councils should be better rated than those that are appointed. By looking at local government areas that differ mainly in terms of whether their council is elected or appointed, we found that local elections are not enough singly to build strong connections between democratically elected local politicians and citizens. Democratic decentralisation in the context of limited financial resources, and limited autonomy vis-à-vis state governors and political bosses, can set elected governments up for failure to the extent that they are viewed as worse performers than their appointed counterparts.


Introduction
Nigeria operates a three-tier federal system made up of the federal government, 36 state governments and 774 local governments.Whereas assumption of the office of the president or state governor is usually by election, assumption of the office of the local government chairperson can be by election or appointment.Elected local government chairpersons emerge from multiparty elections organised by state governments, while appointed local government chairpersons are handpicked solely by state governors from their party loyalists.Among the 774 local government chairpersons in Nigeria, only 202 (26 per cent) are elected while 572 (74 per cent) are appointed.In other words, although local councils are the closest tier of government to the people, in three-quarters of these, citizens have had no input in deciding who leads them at this level. 1 Democracy theorists and pundits (Marcus, Mease and Ottemoeller 2001;Post 2005;Maruatona 2006;Lindberg 2006;Hiley 2006) seem to strongly suggest that where citizens do not have input into deciding their leaders, there might be some form of disconnect leading to a poor perception of the government in question, disengagement and, ultimately, cynicism.This is, however, not a completely settled argument.For instance, citizens' favourable perceptions of some non-democratic or authoritarian governments such as China (Saich 2016) suggest that citizens' non-involvement in the (s)election of political leaders may not always automatically lead to disaffection.The jury is thus still out on the link between elections and citizen perception of government.
The structure of local government in Nigeria provides a useful context for further investigating this link.With elected and selected councils sharing borders in many places, it was possible to set up a comparative study that examined if and how a local council's democratic status influenced citizens' perception of a political administration.By democratic status we refer to whether a local council is elected or appointed.Existing studies suggest that elected councils should be more responsive to citizens and perform duties in a way that pleases the electorate, in order to get re-elected or get elected to higher positions (Inyang and Abraham 2013;Enejo and Isa 2014).This suggests that appointed councils should be more responsive to the state governors that appoint them, and should seek to please them so as to retain their seat or get selected for higher positions.Accordingly, we hypothesised that an elected local government council would be perceived more positively by the electorate than a selected one because, being closer to the people and being a product of their choice, an elected council should not only initiate programmes that resonate with the aspirations of the electorate but also do so in ways that involve the contributions of the citizens.
We wove our investigation around the provision of security by local councils because not only is security a major issue in Nigeria, but it is also an issue that very often affects how Nigerian citizens perceive the effectiveness of government.For instance, observers pointed out that security was a major reason that the government of President Goodluck Jonathan lost re-election in 2015 (Owen and Usman 2015;Zane 2015).Insurgency, herdsmen-related violence, pipeline destruction and oil theft, kidnapping and ransom-taking are some of the problems that make security a topical public agenda in Nigeria.
Although providing security for citizens is not one of the functions of the local government specified by the Nigerian Constitution, most local governments provide security in one form or another.In fact, the chairperson of the local government council is regarded as the chief security officer of the local government area (LGA).This is despite the fact that local governments are not permitted to have their police or armed forces or to recruit, train or arm any security personnel.In providing security for its citizens, the local government necessarily works with a variety of actors.These include agencies of state such as the police, the Nigeria Security and Civil Defence Corps (NSCDC) and the State Security Service (SSS).They also include non-state actors, such as traditional rulers and other community leaders, local 'vigilante' groups (groups of volunteers that are found nationwide in Nigeria), and market-women leaders.Thus, providing security for citizens can become a local government's way of providing employment, interacting with and getting closer to the citizens, and possibly building a stronger reputation for delivery.
In the context of the decentralisation of security provision to the lowest tier of government, it is important to pay attention to citizens' evaluation of their local government for a number of reasons.First, decentralisation, whether it takes the form of devolution or de-concentration, is expected to broaden local participation in decision-making (Bardhan and Mookherjee 2006).As a result, local communities should feel a sense of ownership, which can only be examined from the perspectives of citizens.Second, most crimes are experienced locally within the community, and are committed against individuals.Therefore, it is individual citizens that are best positioned to evaluate crime prevention efforts.Third, there is often a disparity between actual crime rate and citizens' perception of their safety and security (Ojebode et al. 2016).So long as the purpose of a study is to understand how safe citizens feel, working with crime rate figures rather than citizens' perception can be misleading.Our study is thus based on citizens' perception of the performance of local governments in providing security, rather than on measures of actual security provision at the local level.
We paired two local government areas: Asa Local Government and Surulere Local Government.These two local governments are separated from one another by an arbitrarily 2 drawn border that cuts across them to create two different official political entities, but both areas are very similar in many ways.They face similar security situations, are identical economically, are both largely rural and agrarian, and their populations are similar in linguistic, cultural and religious terms.However, there is a key difference.While Asa has an elected local council, Surulere's council was appointed by the state governor.If our hypothesis holds true, citizens of Asa should express a more positive perception of their local council than citizens of Surulere.Contrary to our hypothesis, however, our empirical data reveals that in spite of their shared common experiences of insecurity types and levels, citizens of Asa, the elected council, remained deeply disaffected with their local government, while citizens of Surulere, the appointed council, were satisfied with their council's performance.This is despite the fact that Asa's local government actually initiated a new project to deal with security provision, while Surulere did not.Disaffection or satisfaction was not a product of experience of crime, or actual performance by local governments.Rather, it was a product of the approach to crime prevention adopted by the council, which in itself is traceable to the democratic status of the council.Our analysis revealed that in its bid to respond to and impress the electorate, the elected council initiated new security programmes, recruited citizens and placed them on a regular salary but was unable to sustain these programmes.This situation led to disaffection on the part of the citizens.The appointed local government, on the other hand, was committed to pleasing the state government only; it did not set up any new security programmes nor did it recruit any citizens.Its support for existing informal and community-based security arrangements was modest and irregular, but consistent.This was appreciated by citizens who, as a result, expressed a positive perception of the council.The rest of our article draws out this argument in detail.
2 What are African elections worth?
A number of issues with the quality of national and local elections in Africa have been pointed out by different scholars and commentators.Low voter turnout, violence during elections, and arrant rigging have been the recurring features of elections almost across the continent (Agbaje and Adejumobi 2006;Ikpe 2009).Yet, elections are now a regular feature of African politics, and are organised by most types of regimes, even if many of their problems remain (Brown 2001).Ake (1996: 137) has pointed out that despite several rounds of elections, 'authoritarian state structures remain, accountability to the government is weak, and the rule of law is sometimes nominal.More often than not, people are voting without choosing'.More recently, Marcus et al. (2001: 113) describe African elections as the elite's 'mechanism for lending authority to otherwise authoritarian regimes, rather than creating competitive processes for the attainment of power' while Adejumobi (2008: 362) describes them as a fading shadow of democracy which endangers the democratic process itself by leading Africa back to the old order of 'despotic rulership'.
The poor quality of elections leads to loss of accountability and legitimacy.Lindberg (2003) studied the electoral practices and tactics of Ghanaian members of parliament and found a resurgence of neo-patrimonial practices as politicians engaged in patron-client relationships to reproduce their political power.This threatens both vertical and horizontal accountability.Growing apathy and cynicism are other consequences of repeated rounds of elections that fail the standard tests of transparency and equality in Africa (Ake 1996;Maruatona 2006).
Yet, African elections, despite their imperfections, matter.Beyond Bratton's (1998) dictum that there can be elections without democracy but there can be no democracy without election, Lindberg's (2006: 143) rigorous study of the significance of African elections shows that 'elections do not have to be free and fair or fully democratic to have democratizing effects'.He goes on to demonstrate that even though they were grossly imperfect and far below standard, there was a strong correlation between frequency of elections and the growth of democratic ideals, as operationalised by Freedom House ratings.Going beyond ordinary correlations, Lindberg (2006) demonstrates the causal pathway that elections in Africa take to produce democratic outcomes.
We imagined that the same principle applies to local governments as well, where repeated local elections might produce cumulative positive democratic outcomes over time, including on accountability and responsiveness.Applying Lindberg's (2006) findings on democratic 'lock-ins' to the context of the current study, we can expect there to be a marked difference in the performance of elected local councils and that of appointed councils.If we can hold other things constant, this difference should be attributable to the conduct of elections and the democratic status of local councils.This now is our methodological challenge.Importantly, we expect this marked difference to be reflected in the way citizens perceive and evaluate the performances of their local government authorities, such that citizens of local councils where there are regular elections will evaluate the local authorities better than those in local councils where council leaders are appointed by the governor.Our two selected cases serve our purpose well here -Asa Local Government has remained an elected local government since Nigeria's return to democracy in 1999, while Surulere has moved back and forth between being elected and being appointed, but has had an appointed status since 2010. 3

Decentralisation and local government security services provision
Provision of security services is a constitutional duty of the federal government, the same government that controls the security and crime prevention apparatuses, such as the SSS and the police.However, crime and insecurity records show that, especially in Africa, the federal government is failing in providing these crucial services in spite of these apparatuses.Thirty-one per cent of nearly half-a-million homicides globally in 2012 were committed in Africa (UNODC 2013), and Nigeria ranked among the top 20 nations with the highest crime rates in the world (Numbeo 2015).The Armed Conflict Location and Event Data Project (ACLED 2014) ranked Nigeria the deadliest country in Africa, with about 29 per cent of all organised armed conflict-related fatalities in 2014.That was the year the country recorded 6,383 deaths largely caused by the insurgency in the North.Property theft, kidnapping and pipeline vandalism are among the other rising causes of insecurity in Nigeria.
However, Nigeria, like many other governments all over the world, has come to the conclusion that state policing alone is incapable of solving or abating insecurity owing to historical, economic, cultural and other limitations (Killingray 1986;Marenin 2009).As a result, the country has been promoting different forms of community-based crime prevention (Ojebode et al. 2016).Since these are activities at the community level, they undoubtedly call for the involvement of the local government.There are security committees within local government councils that are headed by the chairperson and bring together all state security agencies with the members of more informal community-based crime prevention groups, such as local vigilantes, hunters and watchmen, and community leaders.Thus, although providing security services is not one of the duties of the local government as enshrined in the constitution, the Federal Government of Nigeria has devolved that duty to the local government level, designating the local government chairperson as the chief security officer of the local government area.
Literature on decentralisation is indecisive on whether or not devolved service provision is effective.On the one hand, local governments' involvement in the provision of services is believed to broaden access and improve service delivery (Kolehmainen-Aitken 1999;McLean and King 1999;World Bank 2001), enhance accountability (World Bank 2000), minimise violence (World Bank 2001) and empower local communities (Brinkerhoff and Azfar 2006).On the other hand, local government provision of services could create a platform for clientelism, budgetary constraints and bureaucratic bottlenecks (World Bank 2001).The tendency also exists for local council leaders to overreach their mandates in a bid to garner popular support during elections (where councils are democratically elected), creating high expectations among the citizens without sufficient budgetary and technical resources to deliver (Rawnsley 2001).
This inconclusive state of the existing literature necessitates contextspecific investigations such as this one.Within the broad construct of decentralisation and in the specific context of Nigeria, we seek to understand citizens' evaluation of the performance of local councils in the decentralised provision of security services, and also why citizens hold such views.
The constitution of Nigeria mandates local councils to be elected by citizens, but at the same time it also allows state governments to select members of a local council if they believe an election is not feasible.Using this constitutional provision, many state governments do not organise elections and prefer to nominate local councils.This decision is entirely that of state governors, and the reason most often cited for nominating local councils is a lack of funds for organising elections, or a tense atmosphere in the state that can ignite electoral violence.A popular perception is that this happens most often when governors are unsure if their party would win the local councils.In other cases, some state governments are simply unable to organise local government elections owing to resource and logistical constraints.
These different practices allow us to construct a descriptive typology of the democratic status of local councils.When the ruling party selects local councils from among party loyalists, we can think of these Type 1 councils (as seen in Figure 1) as 'state envoys' or representatives of the state government and the ruling party.Their assumption of office does not involve the citizens and they are also the least autonomous type of local council vis-à-vis state governments.This type of local council exists in many states in Nigeria, and constitutes about two thirds of all local governments.When states organise local council elections and the ruling party wins, these are Type 2 councils, which we can call 'dual loyalists' because they have both the ruling party and the local electorate to serve.Quite often, such councils emerge amidst widespread allegations of electoral irregularities, which are made more credible by the fact that the ruling party is empowered by the constitution to constitute the electoral body and appoint the chair of the electoral commission, so that elections can be manipulated to bring their chosen candidates into power.However, these are elected councils and have far more legitimacy and possibly also greater autonomy vis-à-vis state governments than nominated councils.
Type 3 councils -selected by the opposition -are possible theoretically but not in practice.It is usually only the ruling party that is able to select local councils.If this does ever happen, it would be a token by the ruling party to pacify the opposition, or a strategy to bring it on board a coalition.On some occasions, the opposition party is able to wrest the control of some local councils from the hold of the party ruling at the state level.When this happens, it is often an expression of the locals' dissatisfaction with the performance of the state government and an invitation to the opposition to prove that it can perform better.There have been very few local councils of this Type 4 variety in Nigeria, and there are none at the current moment.Given how hard this has proven in the history of democratic decentralisation in Nigeria, we can call this category of local politicians 'electoral heroes' for managing to win against the state government.This type of council, when it exists, may enjoy the greatest legitimacy and autonomy vis-à-vis state governments.
Each type of democratic status has implications for the extent to which local councils can exercise autonomy in service delivery.However, given the distribution of councils in Nigeria at the moment, we focus on just Types 1 and 2 -the state envoys and dual loyalists.Restated in the terms of this typology, our hypothesis is that with regard to the performance in providing security services to citizens, dual loyalists would enjoy more positive perception by the electorate than state envoys.The rest of this article checks on the extent to which this is true.

Methodology
Experimental designs come in handy when researchers try to determine if and how a certain cause produces a certain outcome.These designs, which are necessarily comparative, allow the researcher to control the environment of the research, randomly select their subjects and hold some factors constant while manipulating others to determine the effects of these other factors on the outcome (Wimmer and Dominick 2011).
In many study settings, pure experiments are impracticable and evidence of causation has to be based on quasi-experimental designs.One of the key differences between pure and quasi-experiments is the inability of the latter to randomly assign subjects to groups.In other words, researchers have to work with existing groups, not the ones they randomly created.This raises the question of internal validity -a doubt as to the extent to which differences between the groups are solely the outcome of the manipulation of a variable or the presence or absence of an intervention.Yet, carefully constructed quasi-experiments can produce valid causal claims.
With the intention of studying if and how the democratic status of a local council influences the way citizens perceive its performance in providing public goods, specifically security services, we adopted the cross-case spatial comparative case study design.Paying attention to the nuances in the democratic status summed up earlier, we selected two diverse cases: a council appointed by the state governor, and one elected by citizens.We controlled for spatial, ethnic and socio-demographic confounders by choosing contiguous local government councils located in neighbouring states.The two local government councils are Surulere in Oyo State (appointed council) and Asa in Kwara State (elected council).They are separated by the artificial boundary between Oyo and Kwara states.Both local councils have Yoruba-speaking populations that are mostly rural subsistence farmers and traders.Both local government areas also experience the same kind of insecurity problems -mostly property theft and clashes between farmers and herdsmen.In both of them, there are formal crime prevention agencies (the police, the SSS, and the NSCDC) and informal actors, such as vigilantes, hunter-volunteers and other forms of community-driven crime prevention arrangements.
With reference to factors that might affect security provisioning, such as population size and population mix, neighbourhood density, level of urbanisation or industrialisation, presence of state security agencies, and level of unemployment, among others (Ojebode et al. 2016), the two local government areas are identical.According to the 2006 census, Asa has a population of 126,432 (Kwara State Government 2015) while Surulere has 126,692 (Oyo State Government 2015).Both have population densities below the national average: 90.3 per sq km for Asa and 129 per sq km for Surulere against the national average of 200 per sq km.Furthermore, both Kwara and Oyo states are ruled by the same political party, the All Progressive Congress (APC).The main observable difference between the two local government areas is that for one the executive council was elected, while in the other it was appointed.
We began with the collection and analysis of quantitative survey data, a stage designed to show citizens' perception of the performance of their local councils in the provision of security services.The outcome of that analysis called for, and thus guided, further in-depth qualitative investigation in a second round of data collection.This second round focused on understanding why the local councils were perceived as they were by the citizens, and helped us unravel the unexpected manner in which democratic status influenced citizens' perception of security provisioning by leading us to map a causal pathway between status and security provisioning.The integration of both findings led us to a third and final round of data collection, which was also qualitative and was meant for the purpose of validation through in-depth interviews with community leaders and security officials, and through field observations.This sequencing and combination of methods was advantageous.As we show later, the analysis of the first wave of quantitative data collection largely nullified the hypothesis that the elected council would be better perceived by its citizens than appointed councils.The second wave of data collection focused on why this was so, as it tried to establish the link between democratic status and citizens' perception, while the third wave helped in further clarifying this and validating our claims as we discussed our findings with community leaders and security officials.
Although our case and unit of analysis is the local government area and our design is, therefore, a small-N design, our data come from an aggregation of the views of randomly selected individuals.We selected 235 respondents to the questionnaire using a combination of systematic and convenience sampling techniques.Focusing mainly on the headquarters of the local government areas -Iresadu town in Surulere council, and Afon town in Asa council -we selected the first willing adult respondent each from every fifth house with a random start.This gave us 139 males and 96 females; 121 respondents from the Surulere Local Government Area and 114 from the Asa Local Government Area.
As part of the qualitative data collection, we conducted 16 interviews with local government officials, traditional community leaders, members of the Police-Community Relations Committee (PCRC), two officers of the NSCDC, and some community members and leaders.We included security personnel because we wanted an insight into the structure of security services rendered or supported by the local council, especially to be sure these were similar in the two councils; we included community members and leaders to gain additional insight into community perception of the local council.
6 Similar structure, similar experiences of (in)security We found a tripartite structure of security provisioning in each council that comprised government services; community-led and community-driven security services; and private and commercial services.The relationship among these is somewhat nondescript and required our attention.
The first set of security services are provided by the government and involve the police, the SSS, the NSCDC and the local government officials.These agencies are independent of one another: the police, SSS and the NSCDC take orders from their state headquarters located in the state capitals, while the local government officials take orders from the local government chairperson.Together, they form a committee that meets once every month to deliberate on security issues and sometimes to plan joint actions.The committee also includes community leaders.There are no guidelines or standard operating procedures guiding the collaboration between these parties, and the chairperson does not have power of veto.He is more of a coordinator than a chairperson.The main funding for security agencies comes from the federal government through the commanding offices in the state capitals.However, local government councils support the police, SSS and NSCDC materially by donating vehicles (including motorcycles), and sometimes also maintaining these vehicles.
Disagreements sometimes occur among them, as happened during the last general elections when a chairperson deployed certain anti-opposition tactics which were interpreted by the security agencies as being too political and which 'our commanding officers (in the state capital) did not sanction'. 4Such disagreements sometimes fester to the point that the committee members have to assert their independence and their loyalty to their commanding officers in the state capitals, over the wishes of the council chairperson.
The second set of community-driven security services are mainly the activities of neighbourhood or village-based associations led by a community leader, or baálé.These associations or network of associations hire watchmen who work mainly nights and pay them from contributions by members.They may hire members of a registered and recognised national association, the Vigilante Group of Nigeria.Some communities hire members of hunters' associations -those who professionally hunt for game, mostly in rural areas -that provide protection to the community for a fee.In many cases, vigilante members simply volunteer 'without receiving a dime' from the local government. 5However, there were indications that the two local councils not only involve community leaders in the security committee but also support hunters and vigilante groups financially and in kind.They do this using different formats, to which we will return shortly.
The third set of services comprise private and commercial security arrangements by banks and a few rich people who can afford these.This category is independent of the government and community-based security services, and are therefore not included in this analysis.
The three types of security arrangements -government, communitybased and private security -are unevenly distributed across the local government areas.Whereas all three are clearly visible in the local government headquarters, the presence of government and private services wanes as one moves from the headquarters to the interior, more rural parts of the jurisdiction, where community-driven services become more prominent.This has a key implication for our understanding of decentralisation.Rather than being a process that brings government and administrative decisions closer to the grass roots, decentralisation has become the process of multiplying little centres or nuclei rather than of scattering the centre and moving government closer to the citizens.As noted by Kauzya (2007), 'if decentralization has to be successful it needs to be conceived as the transfer of power and authority to the people and not only to local governments'.Local government headquarters have become little state capitals where facilities and government presence are concentrated, and they present a sharp contrast to the rest of the local government areas that lie just a few kilometres away.
The similarities between the two local governments go beyond the structure of security services that exists.There are also similarities in citizens' experience of crime.A Chi-square test 6 for independence (with Yates Continuity Correction) was conducted to test whether citizens in Surulere, the appointed local government council, experienced more crime than those in Asa, an elected local government council.We asked citizens about whether or not they have experienced a serious crime.The results in Table 1 indicated no significant association between status of local government and experience of crime.To be significant, the p-value needs to be .05or smaller.In this case the value of .466 is larger than the value of .05,so we concluded that our result is not significant.The effect size (phi = -.058) is also considered a very small effect using Cohen's (1988) criteria of .10 for a small effect, .30for a medium effect and .50 for a large effect.In addition to comparing the local councils on the frequency of crime experienced by citizens, we also compared them on the type of crime experienced.This is important because types of crime experienced can also influence perception of safety and how citizens judge the performance of government.Here too we found no significant differences across the two local government areas.As Table 2 shows, a 2 by 7 Chi-square test for independence shows that there was no significant association between the democratic status of local government and types of crime experienced by its citizens.The effect size (Cramer's V = .202)is also considered a very small effect.
The two local government areas are similar in terms of the structure of security service provision at the local level, and citizens' experience of crime.This adds to the earlier similarities we pointed out in terms of culture, demographics and economics across the two local government areas to show that they are largely similar units.And yet, the perception of their citizens of the performances of their local government councils is not the same, despite the fact that the council that is perceived negatively has actually initiated security projects, while the council perceived positively has not.We look at and explain this curious finding in the next section.

Different levels of satisfaction with local government
We asked respondents to state their level of satisfaction with the security services provided by their local governments.Table 3 shows the result.An independent-samples t-test 7 was conducted to compare the level of satisfaction with security services of those whose local government leaders were appointed and those whose leaders were elected.The p-value results show that there was a statistically significant difference (p<0.05) in satisfaction here -those whose local government leaders were appointed were more satisfied with security services than those whose local government leaders were elected.The magnitude of the difference in the means is -0.62, and about 10 per cent of the variance in satisfaction with security services is explained by the groups (appointed and elected).
The contrast in the perceived performance of both councils is starker in the interview data.Interviewees in Asa Local Government Area (the elected council) could hardly find a kind word to describe the security services of their local government council.Recurrent references were made to the recruitment of the vigilantes and the local government's inability to pay them.Some claimed that even when the vigilantes were being paid, they were deployed to guard government property not citizens.An interviewee 8 in Asa (elected) Local Government Area stated why the local government was unresponsive: The local government authority has done nothing to ease the pains of the security threats… They have not performed to the expectations in other aspects too.In my opinion, they have good intentions but are being thwarted by financial constraints.Joint Account with the state government has made progress at local government impossible. 9I am an insider myself.The local government is paralysed; they can't even pay the salary of their workers.
On the other hand, interviewees in Surulere Local Government Area (an appointed council) were satisfied with the performance of the local government council in terms of security provision.An interviewee 10 in Surulere Local Government Area (appointed) said: The Local Government Authority has been a tremendous supporter of the Security Agents.They assist with money and other necessary gadgets.The police and vigilantes have benefited from this good gesture.They are doing very well.
In spite of the shared common experiences of insecurity types and levels across the two local government areas, citizens of Asa, the elected council, remained deeply disaffected while citizens of Surulere, the appointed council, remained satisfied with their council's performance.
The difference in the level of satisfaction could not have been a product of experience of crime or a result of any cultural, economic or demographic differences, since these were held constant.As the next section shows, it is rather a product of the style of security service provision of the local government councils, which is itself traceable to their democratic status.Why would the citizens in a local government area perceive the security services of its local council poorly while another next to it perceives those of its local council in much better light when these two groups of citizens have similar experiences of insecurity and similar security provision structure, among other germane similarities?Our analysis shows that the reason for this disparity in perception was not the security situation or the performance of the local government authorities but rather in the approaches adopted by each council in providing security.We discovered that these approaches are a reflection of the democratic statuses of these local government authorities.The two different approaches are presented graphically in Figure 2.
Both local governments adopted a formal approach to working with formal agencies such as the police, SSS and the NSCDC.There are regular monthly meetings, with minutes of meetings and other formal procedures.However, they have markedly different approaches to working with informal security providers, which constitute a significant part of security provision in both areas.Asa Local Government Council, an elected council, adopted a formal approach in providing security services with the traditional informal security service providers -the community leaders and associations, vigilante groups and local hunters -by starting a major project with these groups.They were recruited, trained, kitted out and put on a monthly salary of N7,000 (less than US$20).However, the local government could not sustain this and the arrangement failed, creating a feeling of disaffection on the part of the citizens.On the other hand, Surulere Local Government Council, an appointed council, adopted an informal approach to working with informal partners.By implication, the appointed council did not have a commitment to regularly support or sponsor the informal agents.If there were funds, the council supported the agents, if funds were unavailable, the council did not support them.Why did the two local governments adopt such different approaches?

Asa
We probed further to understand why the elected council in Asa would embark on the formalisation of an informal but widespread and reportedly effective security arrangement (Ojebode et al. 2016).
Recruiting and kitting up local vigilante members and hunters, it seems, has apparent strategic and intrinsic value for security provisioning.Being locals, they have a thorough knowledge of the geography of the area and can therefore monitor movements and follow trails.Not only this, they know the residents in the communities and have a better idea of likely offenders than the police would have, and they are also known by the community, and so are most likely to abstain from conduct unbecoming of a community member -such as extortion or victimisation.Selecting them and training them was thus strategic.Closely related to this is the fact that recruiting and kitting out locals was also a way of providing employment and reducing the incidence of crime.
There are extrinsic values of this strategy as well.The council strategically chose a method that would have the greatest impact on the psyche of the locals: official uniform, training (including drills and march-pasts) and kits.In a seminar paper presented on the non-verbal significance of official uniforms, Deinde (2013) explained that the official uniform in colonial and postcolonial Africa was more than mere clothing or, as it is in the West, an expression of state presence.It was also an expression of the wearer's worth and importance.To kit a person out was, therefore, to give them legitimate state presence as well as enhance her or his worth in the community.This, according to him, explained the awe with which colonial police officers were treated and the widespread cases of corruption against them even at that time.
At the beginning, our interviewees observed, the programme attracted the attention and commendation of communities for its ingenuity and its immediate impact, not just on security, but also on employment.
Everybody was happy.Hunters and many of these other jobless people were bringing something home at the end of every month.They had uniform and salary.Like soldiers.And their families were happy too and they were proud.And people said this chairman had good ideas and he was like a saviour.We said he made (electoral) promises and we voted for him, and now he is fulfilling them. 11 However, the chairperson could not sustain the programme beyond the first 18 months.A community leader who also worked in the local government informed us that funds coming in from the federal government through the state reduced, and internally-generated revenue did not increase.Soon, the chairperson stopped paying the recruited and kitted-out locals.
When our salaries stopped coming, (there was) nothing to show at home at the end of the month.What is uniform?What is boots?Is that food?People even made jest of us; uniform without salary. 12  The economic and social aspects of this funding discontinuation was negatively felt by both the affected workers, their dependants and others in the local government area.The eventual outcome was palpable disaffection against the local government authority.
A simple depiction as the one in Figure 3 risks over-simplifying a complex situation, yet it is invaluable in laying out the basic skeleton of a complex problem.Figure 3 summarises the likely reasons the elected council may be perceived negatively by the electorate, in this case with reference to their performance in providing security services.
An elected local government chairperson is a dual loyalist: he has the electorate to serve but also has party stalwarts to please.Figure 3 shows that the dual loyalist faces pressure from two separate directions.At the top of the list of these loyalists are those influential people, called 'godfathers', who dictate the distribution of party tickets to contest elections and can influence votes and victory.Edigin (2010) points out that large resources are expended to please and service these godfathers or 'kingmakers'.Those who refuse to serve the godfathers after their electoral victory are 'subjected to indecent manipulations' and risk being impeached (Albert 2005: 105).Edigin (2010) cites examples of such impeachments, which godfathers can affect either through legislative means, which they also control in many cases, or through more informal means, such as gangs that are used to create violence and other forms of crisis in the jurisdiction of their estranged 'godsons' (Ellis 2008;Olarinmoye 2008).We do not have firm evidence of how many resources went into pleasing the godfathers in our two cases, but most analyses suggest that Nigerian politicians expend substantial financial resources to service godfathers.In fact, Albert (2005: 103) explains that while there are powerful people all over the world who are capable of influencing electoral victory, the Nigerian political godfather is unique in that he has 'turned politics into a money-making business… an economic investment that must yield superlative dividends by all means'.The second source of pressure comes from the electorate and their expectations of effective delivery, based on promises made during election speeches, rallies and campaigns.To meet these, the elected council chairperson creates big projects, in this case security projects.However, as funds eventually reduce but demands from godfathers and party bosses do not, projects are abandoned.People lose jobs and the community at large witnesses other obvious signs of the local government moving away from projects and electoral promises, and as disaffection sets in, the chairperson is seen as yet another deceitful politician by citizens.

Surulere
Why would a state envoy, that is, the appointed local council chairperson, neither adopt a formal approach to security nor initiate a new security project?The chairperson of Surulere local government council did not initiate a project; he depended on instructions from the state government for every project.He had no independence and could not recruit workers for a project except with the approval of the state governor.Since he did not campaign for elections, he did not make promises and therefore, there is little or no expectation from the electorate.Given the ubiquity of godfather-ism in Nigerian politics, as discussed previously, we suggest that he is faced with pressures from godfathers as well, but that is the only set of pressures he faces since there is no electorate in this situation.
Our interviewees from Surulere expressed satisfaction with the performance of the local government in security provision.A leader of the local watchmen (hunters) whom we interviewed 13 claimed that the chairperson: supports us very well.He donated motorcycles to us and also gives us money from time to time… not every month.We don't get money every month since we are not civil servants.Sometimes the money is much; sometimes the money is small.We all know that things are hard in this country.
It is important to note that the motorcycles were regarded as donations rather than entitlements.Donations evoke a sense of gratitude; entitlements do not.
Another interviewee commended the chairperson of Surulere for occasionally attending their meetings or sending his representatives.This was a sign that the chairperson recognised and respected them.Others spoke of his 'closeness' to the traditional rulers which is interpreted as a sign of his humility and respect for tradition and culture: 'He does not pass on this main road without stopping to greet our leader'.That these simple and generally common gestures are commended shows that the citizens did not expect these from the appointed chairperson who, in their view, is not under any obligation to make such gestures.
Occasionally, when specific needs arise, the leaders of the informal security agents approach the local government chairperson in Surulere for financial assistance.One such instance was when the vigilante group wanted to buy uniforms.The chairperson donated some money for this.At another time, he donated money for repairing their motorcycles.At other times, he simply invited them and gave them some money for their operations and logistics, such as when the state governor visits the local area and all informal agents are mobilised and put on high alert.
The spontaneity of such acts, and the fact that they were unexpected, left the informal agents deeply impressed.
From the perspectives of the citizens of Asa and Surulere, one chairperson represented unmet expectations and unfulfilled promises, while the other represented unexpected gestures.The former was thus regarded with disaffection while the latter was positively rated.It seems that democratic status and its attendant politics matters in citizens' perception of the performance of local councils, and that the pathway works through two particular factors -whether citizens have expectations of delivery from their local governments, and whether there are enough resources available to elected leaders to meet the various pressures that they must operate under.
9 Conclusion: is being elected a disadvantage?
The evidence presented in this article shows that local elections in a context of party control, godfather-ism, and a lack of financial resources and autonomy sets elected local councils up for failure, and disaffection from their constituents.Elections lead to a sense of entitlement by citizens, a particular strength of democracy, but this needs to be accompanied by an ability to deliver on electoral promises and on the particular expressed needs of the electorate.When promises are not met by delivery, disaffection results.Where no promises are made, any delivery or gesture by leaders is unexpected, and therefore, considered generous.In either case, the imperatives of democracy and the needs of citizens are not well served.

Figure
Figure 1 A typology of democratic statuses

Figure 2
Figure 2 Approaches adopted by elected and appointed local governments with formal and informal agencies

Figure 3
Figure 3 Causal pathway from elected democratic status to poor perception of performance

Figure 4 Causal pathway from appointed democratic status to good perception of performance
findings on the significance of African elections and on democratic 'lock-ins', our findings suggest that local government elections that operate under political and financial constraints may actually create a constituency against further democratic decentralisation.If community members perceive the performance of those democratically entrusted with leadership positions as being worse than those appointed to these positions by party bosses, then it is possible that support for elected local governments may reduce over time.Making democratic decentralisation work requires a legal and constitutional framework that protects elected council chairpersons from the stranglehold of political 'godfathers', and provides them with some measure of fiscal independence and autonomy vis-à-vis state governors to provide services.We are grateful to the Partnership for African Social and Governance Research (PASGR) for a grant for this research, and to two anonymous reviewers and Shandana Khan Mohmand at the Institute of Development Studies for detailed comments on our article.1 An appointed local council is made up of a caretaker committee headed by the chairperson: all the members of this committee and the chairperson are appointed by the state governor.An elected council is headed by an executive chairperson who chooses his/her own cabinet. 2 Like most African borders, these borders are not coterminous with language, ethnicity or cultural boundaries.They are arbitrary.3 About four months after our data collection, the Kwara State governor dissolved the elected council of Asa Local Government, thereby turning the local government into an appointed one.