Introduction

Since the late 1980s, there has been a great deal of international support for decentralised governance to offer a comprehensive range of benefits, from the increased efficiency and effectiveness of public services provision to increased state legitimacy, stability and support, and reduced corruption (Crook and Manor, 1998; Gaynor, 2016). These forms of decentralisation have hoped to respond to the institutional frailties and development challenges in developing countries in Asia, Africa and Latin America. These reforms are further intended to break up over-centralised governments (Arkorful et al., 2021). This is largely due to the widespread international support for decentralised reforms adopted in over 80% of developing countries to date (Gaynor, 2014). Furthermore, since decentralised governance places emphasis on the link between local governments and community participation, community participation programmes are launched to empower citizens to become actively involved, thus the governance will be more democratic and more effective (Irvin and Stansbury, 2004; Pius Kulipossa, 2004).

Equally, it is contended that sometimes states introduce local government reforms according to their old ideology (Staniland et al., 2020). In Pakistan, military dictators such as General Ayoub and General Zia-ul-Haq saw the advantage of using local governments to conceal a highly authoritarian regime (Abbasi, 2006; International Crisis Group, 2004, p. 4).

In agreement with some commentators (Ishii, 2017; Siebers et al., 2019), we argue that the local governments under ruling non-democratic regimes alienate and frustrate the citizens, engender mistrust and strengthen local political elites.

The purpose of this article is to explore the dangers of decentralisation under non-democratic historical institutions in Pakistan. The key question implicit in this aim is ‘What are the potential threats of decentralisation introduced by the military regimes on public participation?’

This paper addresses a gap in the research literature on the contextual nature of a state based on historical institutional frailties and how local government reforms can pose a danger to community empowerment and public participation. This empirical study used the lens of historical institutionalism to explore the nature of the intentions of a military establishment in using decentralisation to undermine the expected positive benefits to ordinary citizens.

The article is structured as follows. First, there will be a brief review of the literature on decentralisation as a premise for establishing democracy through citizen participation, then, using the theoretical lens of historical institutionalism, the authors will explore how successive military regimes in Pakistan have used seemingly democratic institutions to control the internal and external challenges.

The political context of Pakistan is used as an epistemology to view the motives and results of the various decentralisation reforms introduced by a dominant military regime.

Literature review

The major premise of decentralisation reforms is that they bring governments closer to citizens through the mechanism of citizen participation by introducing accountability into local governance (Botes and Rensburg, 2000; Barter, 2008). Decentralisation has been defined as the meaningful devolution of central powers to local units of governance, which are accessible by and accountable to the local population (Blair, 2000; Vergara, 2015).

The focus here is on ordinary citizens being invited to become involved in local-level governance activities in local development works (see Table 1).

Table 1 Participatory mechanisms.
Full size table

The expectations from decentralisation might be inspired by the assumption that “decentralisation offers opportunities for greater personal participation in the actual business of governing and to create a democratic climate” (Smith, 1985, p. 20). These expectations can only be realised from the institutions formed by the state that are democratic and can offer sufficient opportunities to the common masses to participate in the development process (Reedy et al., 2020). The literature supports the notion that decentralisation’s contribution to community participation is associated with the contextual environment (Gaynor, 2014; 2016; Ishii, 2017; Reedy et al., 2020). Some commentators (Botes and Rensburg, 2000; Buček and Smith, 2000; Ishii, 2017) have expressed concerns that in some countries, public participation through a decentralised system is challenged by various adverse factors, such as the historically authoritarian nature of the centralised state, a culture of traditional feudal lords (local elitism), and unequal societal structures. The local governments have been used simply as a tool for controlling the emerging empowerment of the local leadership. In such contexts, the impacts of decentralisation reforms must be carefully analysed, otherwise it can be an unrealistic burden of expectations regarding its ability to transform whole societies dominated by military regimes or patronage politics (Rondinelli, 1991; Crook and Manor, 1998).

The first challenge to decentralisation and public participation is that authoritarian regimes often introduce a local government system to gain support from the new political groups and create a façade of reforms for their donor audiences and foreign investors in order to integrate themselves into global markets (Ribot et al., 2006; Gandhi and Lust-Okar, 2009; Kyriacou et al., 2017; Aslam, 2019). Such reforms are often introduced in a bottom-up approach, which is merely a nod at the devolution of power and therefore do not necessarily empower the common masses (Abbasi, 2006; Ribot et al., 2006; Mohmand and Cheema, 2007).

The second aspect is that local government reforms are simply smokescreens obscuring the intention of a power retention channel for an authoritarian regime. In Pakistan, the newly elected nazims (district mayors) did their best to respond to the central authorities and not to the sub-national local governments. Furthermore, they offered their loyalty to the central military government (Ribot et al., 2006; Mohmand and Cheema, 2007; Aslam, 2019). Thus, a local government system offers a dictator a wide range of mediums for negotiation with various groups in developing countries so that they can co-opt members of the popular opposition parties and identify them as a loyal clientele (Aslam, 2019, p. 127). In this way, authoritarian regimes can use decentralisation mechanisms to recruit new political support to legitimise their non-democratic regimes so that their military rule can grow under the guise of local democracy (Gandhi and Lust-Okar, 2009; Aslam, 2019). For example, in the recent case of Pakistan, the Devolution Power Plan 2000 introduced by the military government boasted of trickling power down to the lower levels, whereas in reality those powers were simply shifted from the provinces to the local districts. Furthermore, as the crucial units from which the marginalised people were supposed to get relief, the union councils remained neglected (Abbasi, 2006; Mohmand and Cheema, 2007).

A recent trend has appeared that authoritarianism and despotism can be adapted to present a benign face to the population and change its appearance so as not to be seen or felt to be arbitrary (Urbinati, 2007, p. 68). In this way, authoritarian governments can continue to launch regime-supportive institutions which do not appear to be regime-subversive: “an authoritarian world without ambivalence seems to be an authoritarian illusion … if dictators wish to reap the fruits of stability and governance from their orchards of political institutions” (Schedler, 2009, p.15). The dangers of such decentralisation are that the non-democratic institution of local government can prevent civic society from rallying against centralised authoritarian regimes and the ordinary citizens are pushed further away from having any involvement in state affairs (Sudhipongpracha and Wongpredee, 2016; Ali, 2022).

The second challenge to militarily sponsored decentralisation reforms is that the local elites notoriously capture these local powers. This capture of local power has far-reaching negative impacts on the local population. For example, it might not only strengthen already dominant feudal lords but also transfer social conflict to the local level where there is greater political inequality (Prud’homme, 1995; Goldfrank, 2002). In this way the local political powerholders can continue to form a barrier between the state and the ordinary citizens. This is what has been happening in Pakistan where the local political powerholders became the only conduit between the citizens and the central state authorities (Cheema et al., 2005). As a result of such a barrier being imposed by the dominant local elites, there is a gap between the state and the citizens of Pakistan (Ali, 2022; 2020; Lall, 2012).

The recurrence of such situations creates more networks of patronage and strengthens the power of the feudal lords or a local elite culture by getting unconditional support from the downtrodden underprivileged masses. Clausen (2020), for example, found that the Yemeni authorities had introduced local government reforms in 2001 and 2006 simply to promote the regime’s internal patronage network and external legitimacy. The danger of this practice is the possible rise of ‘environment despotism’ in which the common public will have to adjust to fit the corrupt activities of authoritarian institutions. In this regard, Mill’s view (cited in Urbinati, 2007, p. 68) was that the nature of authoritarianism will permeate throughout social practices, which will appear to be based on consent while denying individual demands. And further, such decentralisation reforms can also create a localised client class which, as observed above, acts as a conduit between the underprivileged masses and the non-representative central authorities of the state (Cheema et al., 2005). For example, it has been shown that decentralisation reforms are often introduced into neo-patrimonial contexts, which results in a client/patron relationship in which the local elites usually capture the local power (Kakumba, 2010; Gaynor, 2016; Clausen, 2020). The danger is therefore that decentralisation creates more inequality within a state.

For instance, in the case of the Yemeni state, the intended decentralisation was hampered by the structure of the institutions of the state which consisted of various networks of the ruling elite involving both internal and external actors. The local power was mainly captured by the local elite and tribal leaders who, in many localities, were the primary authorities providing limited public services (Clausen, 2020, p. 122). Yemen’s social structure is similar to that of Pakistan’s local tribal leaders because the unequal elite class raises obstacles to the local government’s capacity to produce the desired outcomes.

It seems plausible that a mechanism of decentralisation during authoritarian regimes maintains the practices of a patron/client structure (Barter, 2008). Such decentralisations in the case of Pakistan did not occur between citizens and the state. Instead, the local power was transferred from central authorities (patrons) to clients (local elites) (Mohmand and Cheema, 2007). Some scholars (Ribot et al., 2006; Busygina et al., 2018) have aptly argued that most decentralisation reforms do not materialise in authoritarian regimes and often such failures are linked to de-concentrating the transfer of powers and resources to central governments. The purpose of these reforms is purely to “enact a theatrical image of reforms for their donor audiences” (Ribot et al., 2006, p. 4) and to try to integrate themselves into the global economy (Busygina et al., 2018).

In such a situation, fiscal decentralisation is severely challenged when a government faces an underdeveloped political culture, regional inequalities and bad governance (Muhammad, 2004; Kyriacou et al., 2017). Even though financial departments are run by public officials rather than elected mayors, they nevertheless benefit from some revenue through their own taxes and devolved fiscal autonomy (Martinez-Vazquez and McNab, 1997). Given the absence of democratic governments in these regimes, it can be safely assumed that these reforms are some form of deconcentration of devolution (Martinez-Vazquez and McNab, 1997). In Pakistan, the decentralisation of 2001 happened at the sub-national government level instead of powers being trickled down from the central authorities. Thus, the local governance institutions, including elected officials or permanent staff, have often been managed by authoritarian regimes. As far as bureaucracy is concerned, authoritarian regimes are constantly looking for opportunities to use shortcuts to achieve their targets (Muhammad, 2004). Muhammad (2004) concluded—after carrying out a detailed case study of urban councils in Punjab, Pakistan—that the equal distribution of expenditure under a non-democratic government is impossible, and that meaningful fiscal decentralisation in Pakistan is a mere slogan because financial devolution cannot be achieved under a non-democratic government.

Plessing (2017, p. 74) stated that another challenge is that participatory mechanisms do not work in countries such as South Africa, where elites hold power and influence over the meaning and the resources of implementing a policy of participation and other development projects. Evidence from South Africa showed that decentralisation does not necessarily promote public participation because of the highly unequal contexts of the state institutions. For this reason, there the marginalised communities chose specific forms of participation such as protests and insurgency to get their voices heard (Plessing, 2017, p. 84).

Gaynor (2014) studied the re-introduction of decentralisation in Burundi in 2005 and concluded that despite considerable financial investment in the local government system, the predatory nature of the central authorities enabled the previous neo-patrimonial state to continue to impose and even expand its power. The decentralisation failed to empower the marginalised communities because Burundi’s achievement of independence in 1962 had been plagued by institutionalised corruption, social exclusion and a total lack of accountability (Gaynor 2014, p. 204). Similarly, Hadiz (2004) reported that in Indonesia there were institutional frailties, such as predatory elements of an unequal class structure, political gangsterism and predatory interests of rulers who maintained a centralised system of patronage under the state leadership of the Soeharto regime. In the case of institutional weakness, decentralisation does not always work in the way it is supposed to (Hadiz, 2004).

There can be further negative consequences of underdeveloped political institutions, which could potentially weaken the democratic processes in a civic society (Sudhipongpracha and Wongpredee, 2016). These negative implicaitons can be that “once a single institution is formed, its original attributes tend to persist over time” (Heo et al., 2021, p. 2). We therefore concur with Sudhipongpracha and Wongpredee (2016) and Marriott (2010, p. 37) that decentralisation reforms introduced in many developing countries are merely reflections of their protracted struggle between central and local elites and will have continuing or constraining influences over the policy or other institutions into the future.

We therefore go further to argue that the reforms introduced by authoritarian regimes not only weaken the democratic elements in a civic society, but also lead to inferior governance and stark disparities amongst society (Sudhipongpracha and Wongpredee, 2016; Ribot et al., 2006; Kyriacou et al., 2017).

The context of Pakistan and case selection

Historically, the military dictators of Pakistan showed a keen interest in introducing local government reforms with the purpose of creating a loyal cadre of local elites to support the continuation of their authoritarian regimes—the favoured policy of the various military governments in Pakistan (Mohmand and Cheema, 2007; Aslam, 2019). For example, General Ayub, who imposed the first martial law in Pakistan in 1958, introduced a local government system which was not, in any real sense, fully representative of the government because it could not truly represent either the culture or the needs of the local people (Mahmood, 2000; Ziring, 2003). This was the first historical development of a local government structure, which held out incentives for the military general running the country. The second historical critical juncture in the development of local government happened when General Zia-ul-Haq revived the idea of local government through the promulgation of Local Government Ordinances (LGOs) in 1979 in four provinces of Pakistan (Muhammad, 2004; Abbasi, 2006). Finally, on 14 August 2001, General Musharraf introduced local government reforms throughout Pakistan, including Sindh province, and the marginalised citizens had no choice but to adjust to fit in with the rules of corrupt local institutions supported by the authoritarian regime (Rees and Hossain, 2010). Issues of accountability, elite control and social exclusion have been and still are rampant in Pakistan (Mohmand and Cheema, 2007; Aslam, 2019).

There is evidence of such decentralisation reforms in case studies carried out by Khan and Anjum (2013), who reported that the Citizen Community Boards (CCBs) continued to be less of a local priority for government officials in Pakistan. The main reasons for this were reported to be dishonesty and the misuse of power by local elites. Khan and Anjum (2013) compiled 94 detailed case studies based on interviews and concluded that only the rich and influential were likely to be privileged to participate in and thus to dominate the decision-making process in Pakistan. Kurosaki (2006) surveyed 42 union councils in the district of Hafizabad and found that the quality of the leadership and the establishment of appropriate rules in the CCBs were key determinants of the success of the CCB initiative. Kurosaki went on to emphasise the importance of capacity building in CCBs and the local community.

As far the publicly elected local governments are concerned, the local government system in Sindh was introduced by the government of the Pakistan People’s Party (PPP) in 2013 and implemented in 2016. It is regarded as the first time in Pakistan’s history that elected local governments had been introduced in Sindh (Rid Murtaza, 2018). Notably, the publicly elected local governments of 2013 did not mandate organisations below union councils, which often did not represent communities. For example, Ali (2020, p. 240) found that “in both periods—the local governments of 2001–2009 and 2013—the public was kept away from the decision-making process”. Thus, the marginalised communities were ignored by not introducing an effective mechanism for CCBs.

Methods

Research settings

Despite limitations of time and resources, we sought to represent in this study the urban and rural populations of Sindh, the second largest province of Pakistan. Accordingly, Larkana as an urban district and the rural district of Kambar Shahdadkot in Sindh were selected as the two case-study locations for this research. The province of Sindh is in the southern part of Pakistan and its capital city of Karachi is Pakistan’s chief port and the largest city dominating its coastline (Ansari, 2015). Using Sindh as the location of two diverse case studies provided a unique opportunity to illustrate the issues addressed in this paper.

Sample

Bryman (2016) suggested that anyone for whom the topic is relevant can logically be an appropriate participant in any study. The use of qualitative sampling here was to gain depth and uniqueness rather than breadth because even a small sample of a population can nevertheless provide a detailed account of the issue under examination (Fink, 2003). The representatives who participated in this study were purposefully selected to reflect the salient characteristics of the relevant populations in the two case-study districts (Bryman, 2016; Neuman, 2015). Some topics do not require participants of a particular kind, so there was little if any restriction on who might participate. The logic of whether to choose probability or non-probability sampling depends on what a researcher seeks to find out. We adopted purposive sampling for data collection because relevance is important for collecting first-hand accounts of the phenomenon being researched (Neuman, 2015). Our effective sample comprised union councillors (n = 8) of whom six were CCB members, teachers (n = 21) from different levels of the profession, non-profit organisation representatives (n = 14), members of trade unions (n = 3), local business-people (n = 3), labourers (n = 3), lawyers (n = 2), government officials (n = 2), healthcare professionals (n = 2) and university students (n = 6).Footnote 1

Data collection

It is imperative to keep in mind the scope, purpose and time limits of a research project rather than focusing on the forms of data gathering (Neuman, 2015). We believed that an open-ended questionnaire would elicit unlimited numbers of possible answers by the respondents, who could answer in detail and qualify and clarify their responses.

Purposive sampling is appropriate to select unique cases which are especially informative, … we often use purposive sampling to select members of difficult-to-reach [populations], … a researcher will use many different methods to identify the cases because the goal is locating as many cases as possible. (Neuman, 2015, p. 274)

The collection of qualitative responses from a well-designed and validated questionnaire was carried out by trained field assistants recruited in Sindh. Several researchers (Oppenheim, 1992; Bell, 2014; Saunders et al., 2019) have shown that designing an effective open-ended questionnaire can be challenging, so care needs to be taken that it elicits the required data to answer the research questions. In this study, prior to the data collection process, appropriate ethical approval was sought from and granted by the research ethics committee of the University of Bradford in October 2019. After purposeful sampling had been chosen as the recruitment method and ethical approval had been granted, potential participants were approached by the researchers using social media (Facebook, WhatsApp and mobile phones) and local field assistants were contacted and invited to participate in the study. An introductory letter was distributed by the field assistants in which email and mobile contact details were provided to potential respondents in case any of them wanted to contact us. The respondents were given adequate time to consider the information carefully. They were reassured that participation was voluntary and that no pressure or coercion would be applied while they were answering the open-ended questions. In order to maintain the confidentiality of all the respondents, pseudonyms were used to label their responses and no personal information was presented, which might cause them to be identified. After completing the translation of the questionnaire from English into Sindhi (the local language of Sindh and therefore of the respondents), we dispatched the questionnaires containing the seven open-ended questions, the respondents’ information sheets and the informed consent forms to the field assistants. The field assistants successfully distributed 150 sealed questionnaire copies among the targeted respondents on 19 October 2019, and by 30 March 2020, 70 completed questionnaires had been received. On completion of the data collection, we found that about nearly half (70) of the responses were useable. In many studies, field staff often act as gatekeepers, sources of communication between researchers and the local community. Moreover, as locals, field assistants can make use of the research process to encourage debate and transformation within their communities (Turner, 2010; Caretta, 2015). The questions related to the local government reforms of 2001–2009 included: (1) calls from local government officials to attend local council meetings, (2) attendance at information meetings, and (3) any other form of citizen engagement.

Data analysis

In general, data analysis in the social sciences means a search for recurring significant themes. In other words, as Bernard and Ryan (2010) described it, data analysis is the search for patterns and for ideas, which explain why those patterns occur. This is usually achieved by searching for themes. Our thematic analysis involved three stages of coding: open coding, a second pass of coding and final selective coding (Neuman, 2015). We identified themes, issues and categories, which the questions and objectives were designed to elicit (Saunders et al., 2019). Some researchers have suggested that qualitative data should always be manipulated into a quantitative format (Miles and Huberman, 1994). However, it should be noted that in qualitative studies, numbers tend to be ignored eventually. It is the nature of qualitative data that matters. Researchers have to make judgements about the quality of the data, identifying or addressing issues which occur a number of times and consistent sequences of events, which signifies that something significant is happening again and again, so it is necessary to take notice of that quantitative element in order to be able make a judgement regarding the quality of the data (Miles and Huberman, 1994).

Our process of data analysis started with the open coding technique in which we conceptualised the statements, assigned initial labels and identified various emerging issues from each research question (Neuman, 2015). Saunders et al. (2019) suggested that “… identification of themes/issues or categories should be guided by the purpose of [..] research as expressed through [..] research questions and objectives” (Miles and Huberman, 1994, p. 253).

The reason for using the frequency of identified themes in this study was that we could maintain analytical integrity and protect ourselves against possible bias (Miles and Huberman, 1994).

As another precaution for making the analysis rigorous and with as little bias as possible, we introduced a table to record the frequency of significant recurring themes. Table 2 summarises the method and the frequency of recurring themes in the initial stages of thematic analysis.

Table 2 Frequency recording mechanism and results.
Full size table

Findings

The analysis of the participants’ narratives revealed major themes and sub-themes that illustrated their experience of the ongoing attitude of local governments towards public participation in local development in Sindh. The findings were categorised as follows: (1) local communities were kept at a distance from governance activities in order to maintain the traditional feudal class hierarchies; (2) local elites favoured their own areas and families, thus preventing civic involvement; (3) civic knowledge was deliberately not provided; and (4) the assumption that citizens’ participation would be a challenge to the upper-class status quo. Table 3 shows the number of issues identified and summarises their percentages in the findings.

Table 3 The findings from the study.
Full size table

As Table 3 shows, 55% of the respondents reported that public participation was cancelled or that local citizens were kept at a distance from the development activities of the local governments in Sindh. It is not entirely clear why the ruling elites kept the local citizens at a distance. The most recurrent theme in the findings was that the members of the local community were deliberately kept at a distance from the ruling class so that there would be as much alienation as possible.

More marginalisation of the common citizens

Although theoretically the decentralisation reforms were expected to be democratic, the results showed that citizen participation was not even considered. The following comment from one respondent sheds light on this:

They [local government powerholders] never thought about public participation. This is because once the big man sits on the chair, he thinks himself to be superior to the public. There is no worth and value of law in our lovely Pakistan. Thus, there is no-one to ask the elected representatives why they do not allow public participation. (Male teacher, Kambar Shahdadkot)

This respondent stated that the local powerholders never thought of themselves as accountable for implementing public participation, so they treated the ordinary populace as second-class citizens of Sindh. The reason for this was that there was no system of checks and balances. The locally elected representatives were free to act in an authoritarian way over the local citizens.

The local governments are run from the top command of the party. (Male worker in the education department, Larkana)

This respondent stated that even though the local government mayors came into power through a public vote, their style of ruling was like that of the supreme authority—Musharraf, the architect of the local government reforms. The next extract makes this clear:

During the Musharraf period, they never asked for public participation. All decisions were made between themselves at street-corner meetings between the nazims and naib nazims. (Male teacher, Larkana)

The local publicly elected nazims (mayors) and naib nazims (deputy mayors) acted with complete disregard for public involvement and public accountability. One respondent asserted:

Public participation? This question makes me laugh. This Sindh government and local councils are not ready to consider the common man as a human being. So how is it possible that they would invite us to participate? (Male university student, Kambar Shahdadkot)

The reason for the lack of community involvement was that the local officials deliberately kept a distance between the ordinary citizens and state affairs. If the local government authorities hold the supposed local beneficiaries back, these marginalised communities will not be able to initiate CCB meetings and dialogues: the local representatives consistently sought to isolate the common people from the administrative process.

The respondents explained how local communities were estranged deliberately not only during the military regime (2001–2009) but that the same practice had developed in the local civilian governments (2013–2016).

The various decentralisation systems in Pakistan not only destroyed trust in the local and elected representation system of governance but also pushed the voters further away from state affairs. This led to the common citizens of Sindh seeming to have a negative impression of those tiers, the bureaucracy, provincial and the authoritarian governments, which maintained the least interaction. The literature on decentralisation indicates that if the proper central state systems are inappropriate, decentralisation can cause citizen secession, political instability and ethnic divisions (Prud’homme, 1995; Andrews et al., 2008). Crook and Manor (1998) suggested that it is imperative for decentralisation to bring government closer to the local populace so that the electorate can contribute to effective access to the use of resources. It is quite distressing to note that the local elites, with their ulterior motives, behaved in such a way as to estrange the ordinary citizens from the running of state affairs.

Public alienation

Once common citizens have more knowledge about their rights and responsibilities within the local governance process, they can understand the intentions and consequences of public policies and they will tend to protect and promote their own interests in the governance process (Galston, 2007; Hue and Tung-Wen Sun, 2021). In contrast, however, 22% of the respondents in this study reported that the local government officials did not implement policies of community engagement because of their covert intentions to keep common citizens ignorant about civic knowledge. The next comment by a respondent illustrates this:

They [local government officials] did not invite the public due to fear that once the public is involved in local policies, they will happen to know about their rights, and they will be enlightened with new thinking which will jeopardise the luxurious lifestyle of the ruling class. (Male teacher, Larkana)

Another respondent commented on the importance of citizens being involved and their civic knowledge:

If they [local nazims] had invited the underprivileged voters [to participate], the common citizens would be able to know about their rights. Therefore, fearing such a result of public participation, the local authorities stopped inviting the local community. (Local councillor, Larkana)

Allegedly, the ruling clique did not want there to be any demands from the intended deserving beneficiaries. To succeed in making citizens passive, they avoided putting out any information about state affairs. The findings provide an insight into the view that illiterate, uninformed citizens are easier to rule than people with civic knowledge. The next comment makes this clear:

They [powerholders] have ignored the common public completely … one main view of them is that these people of the upper class do not want the public to be active and to escape from poverty and hunger. (Male local government officer, Kambar Shahdadkot)

Poor citizens are therefore more vulnerable to being exploited by their local political elites.

Authoritarianism and increased alienation

Although there is a normative assumption that decentralisation gives citizens an appetite for participation because local government officials are seen to work for the common good, the case of Pakistan shows that these decentralisation reforms were intended solely to support the military rulers.

They [local mayors] had full power. Musharraf strengthened his coup by allocating many powers [to them]. But the poor remained with no access to their basic rights. (Male social worker, Kambar Shahdadkot)

Another respondent commented that the local mayors were given many powers and that that had strengthened the legitimisation of the military regime:

They [the military rulers] never encouraged public participation. First, because they were not publicly invited but were selected to participate in elections. Second, it was included in their manifesto not to include the public in these development activities. (Male teacher, Larkana)

Twelve percent of the respondents believed that the ruling class was afraid of any empowerment and civic knowledge of ordinary citizens. The reason behind that fear was said to be that once the public become empowered and well informed, the local elites might not be able to exploit state power for their own benefit:

The common public was not consulted about participation. This is because those local representatives on the local councils were the sons and nephews of feudal lords. That was how this family came into the system. (Female college student, Kambar Shahdadkot)

This comment shows that the local government was captured by a specific family (the local political elite) in Sindh. This family had previously been in power at the expense of the marginalised population, so now this elite class was desperate to keep the status quo intact: in order to sustain this practice, local elites always involved their own close-knit family members. The following comment makes this clear:

Because the influential elite brought forward their own favourite groups in local government, the poor were kept away from all services. (Labour department officer, Larkana)

The findings set out above show that the decentralisation reforms of 2001–2009 provided another disadvantage for the citizens of Pakistan in that ethnic and family racism was perpetuated by the existing elites instead of benefiting lower-class individuals. The decentralisation prevented social integration by consolidating the division of society into the rich and the poor and favouring the traditional beneficiaries and their close relatives. So even when the local community members realised that the available funds were being allocated on the basis of patronage, they had no choice but to remain detached from taking part in any events involving decision-making (Francis and James, 2003).

Local elites and the legitimisation of the authoritarian regime

The findings showed that public participation was not possible because the local people who came to hold power were wadera, bhotar and rais (local names for the rich and upper-class in Pakistan). In the presence of these elite classes, the common people could not be invited to participate and give their views about the local policy-making process in Sindh. The responses received from the participants showed that the military dictator had used local government power to obtain the loyalty of the local elites. This view is consistent with the findings of Abbasi (2006), who stated that devolution reforms earned a new client political class for supporting the military regime of Musharraf. Questions arise as to how it is possible that such local government reforms, which are often used for the military purpose of decentralising the state ideology, can empower every citizen in Pakistan. The next comment reveals this:

Whenever anybody sits on the public chair, they start thinking that they are supermen in their authority. Therefore, they did not invite public participation in any service. (Local government officer, Larkana)

This comment indicates that once the local elites (whether sardar, wadera, pir or mir) get hold of power, they do not care about public participation. This suggests that decentralisation has been used for the personal gain of the local government administrators. Once the local politicians of Sindh captured power, they had no consideration for the downtrodden local electorate, as the next comment shows:

Once they got the chair of power, they forgot about the public and whether there was any common public or not. (Male self-employed, Larkana)

Although in practice the military of Pakistan are the sovereign authority of the state, constitutionally, Pakistan is an Islamic Republic.

Discussion

The findings of this study suggest that the mechanism of informing citizens of their rights was not provided to the community in Sindh. In agreement with the theoretical premises of historical institutionalism, the motivating factors behind the arrangements of institutions are critical because they place institutional arrangements on paths, which are then very difficult to alter (Capoccia and Kelemen, 2007).

The key issue here is how did such critical junctures occur in the context of Pakistani institutions. The theoretical answer is that successive military regimes promoted paternalism and authoritarian values with the aim of creating a centralised state (Wilder, 2009; Islam, 2016). The takeover by the military at various times was a critical juncture, which turned Pakistan into a centralised state because, according to the path dependence model, the outcome of a political process is dependent on earlier decisions, which have led down a particular path (McCormick et al., 2019). As a result, the local government institutions set up by the authoritarian regimes acted as controlling tools to support the military governments (Schedler, 2009).

Institutions are coercive structures that mold our acting, thinking and feeling … these institutions shaped the political decisions. (Hague and Harrop, 2004)

Pakistan’s decentralisation reforms not only lacked political will from the central authorities, but also the various authoritarian regimes adapted to the environment and camouflaged the explicitly coercive intentions of a military dictatorship so that it should not be seen as arbitrary (Urbinati, 2007).

As a result of these local government reforms, the marginalised citizens in Sindh were clearly kept deprived of any information about state affairs. The local governments of non-democratic states create even more distance between ordinary citizens and the state.

Thus, the non-democratic state’s local governance strengthened negative gatekeepers (the local powerholders) by enabling them to maintain a firmer hold over the marginalised citizens. Such a decentralisation which benefits a centralised and non-democratic state not only alienates the local communities but also damages their trust, hope, and sense of belonging to the state.

It appears, despite lofty justifications, that decentralisation is laden with the promises of greater community participation but that most local government reforms in Pakistan have been captured by the local elites and, as a result, local power ends up being strengthened without any transfer of power to local people (Agrawal, 1999).

Thus, for local democratic governance to exist, a democratic central state is a pre-requisite for positive results from a decentralised system. This is because without the strong will, commitment and capability of an authoritarian government, the role of local governments might increase the alienation between the ordinary citizens and the state, and result in the loss of resources (Gaynor, 2014; Smoke, 2015).

Conclusion

The research clearly supports the notion that authoritarian regimes launch local government reforms for the purpose of power retention through centralised service delivery, recruiting the clientele class of local feudal lords in order to sustain and strengthen the military regime. These tendencies of recentralisation in service delivery and budgeting push citizens away from the state. Thus, citizens’ interaction with the Pakistani state remained neglected through this mechanism.

We have argued that in order to establish local governments which will have a long enough timespan and space to develop the intended process, with independence from central government influence, there is a need for creating a new supportive civic culture. And local communities must therefore encourage their population to be proactive in choosing leaders and making decisions free of elite control and political interference. Moreover, to keep decentralisation effective in terms of contributing to participatory development, local government and the provision of CCBs should be made mandatory by the constitution.

Local government should not be abolished by any changing government, either military or civilian.

We have also argued that there ought to be a mechanism to minimise political interference and elite control of CCBs, and that the head of the community development department must be a non-political appointment. Similarly, the central government should ensure the availability of adequate staff (dedicated and trained), including engineers (required for the provision of technical support in CCB projects) in community development departments. Local government offices should provide incentives to their employees for their performance in the promotion of CCBs. In addition, local governments should impose penalties on officers of the community development department for not utilising the available CCB funds appropriately.

This paper is limited to evaluating public participation during the various decentralisation reforms introduced in one province of Pakistan by successive authoritarian regimes. We suggest that there is considerable scope for future comparative studies on the impacts of local government reforms introduced by different forms of ruling powers.