Poverty, Power and Local Government in Pakistan

Poverty, Power and Local Government 291 10 Poverty, Power and Local Government in Pakistan Akmal Hussain and Savail Hussain INTRODUCTION In Pakistan, power has been historically constituted within the framework of patron –client relationships: The ruling elite has accessed state resources for arbitrary transfer as patronage to selected individuals for building political support within a structure of dependency. In cases where the resource transfers within such a power structure trickle down to the poor, they merely reinforce dependency rather than counter the processes that systematically perpetuate poverty1 . Therefore, overcoming poverty would involve empowering the poor within countervailing structures of power. In this context, it may be useful to examine the problematique of recent local government reforms in Pakistan for ‘devolution of power’. This chapter examines the dialectic of power between the elite and the poor at both the national and local levels. The implications of this dialectic for efficiency and growth are also examined. The analysis of the power structure at the local level is based on extensive field interviews of local government officials at every tier in the districts of Lahore and Multan. In section-I, we present a historical analysis of the forms in which the patron–client model of governance functioned from the British Raj to the present. In this section, the process of using state resources by ruling elites for building political support in various periods is examined to show how it was associated with the emergence of an economic structure at the national level that tended to perpetuate poverty. In section-II, we present a brief analysis of the dialectic of power between the 292 Akmal Hussain and Savail Hussain elite and the poor. The form of power practised by the elite is counterposed to the form of power implied by the empowerment of the poor. In section-III, the issue of decentralization in the context of empowerment is examined with reference to the lessons learnt from the South Asian experience. In section-IV, the power dynamics of the existing local governments are examined with reference to the issue of efficient service delivery to the poor. Section-V develops an analytical framework for comparing rent seeking and efficiency in local governments and the provincial bureaucracy. Section VI concludes the chapter. POWER, PATRONAGE AND THE STATE Power through patronage has been historically constituted by means of two instruments: (i) The arbitrary transfer of state resources to individuals and factions to create a constituency of dependents who owed loyalty to the Raj (during the colonial period) or personalized loyalty to individual politicians and bureaucrats in the post independence period, and (ii) Discretionary appointments and transfers of personnel within the state sector. In the 19th century, the British colonial government attempted to build a basis of political support, by consolidating the agrarian elite in the areas that later came to constitute Pakistan2 . The British sought the support of the traditional agrarian elite in the province of Sindh by accommodating large landholder families (known as the waderas). In Punjab by contrast, the British formalized the proprietorship over land of the zamindars (large landholders), who had newly emerged from the upper peasant strata following wide spread peasant revolts at the end of the Mughal period.3 In both cases, the colonial government in its early years created a political constituency through establishing patron–client relationships with selected members of the rural elites. In the subsequent decades, the British created new clients amongst the rural elites through offering lucrative appointments in the British Indian Army almost exclusively to the agrarian hierarchy. The most important and far-reaching form of patronage through enrichment of clients was done through the development of canal Poverty, Power and Local Government 293 irrigation and the process of agricultural colonization that accompanied it. From 1885 onwards the British enabled extensive areas to be brought under cultivation through the construction of riverspanning weirs and large networks of perennial canals. These areas that were previously arid waste and had now become arable, were appropriated by the colonial government. Large parts of this newly arable land were transferred as land grants to loyal supporters in the agrarian elites of Punjab and Sindh4 . Additionally, a number of legislative measures were taken by the colonial government to strengthen and protect the position of the loyal rural elites against the operation of market forces. The most important amongst these measures were the Punjab Land Alienation Act, 1900 and the Punjab Pre-Emption Act, 1913, which prohibited transfer of land from land owners to ‘non-agricultural’ classes (Cheema et al. 2001, Ali 1988, Alavi 2001, Pasha 1998). In the post independence period, the patron–client model of governance continued as the bureaucracy in the Ayub government (1958–1969) granted licenses and contracts to favoured individuals in the private sector within a highly regulated economic regime. At the same time lucrative appointments continued to be made in the state sector to establish a domain of patronage for the military–bureaucratic ruling elite. During the 1960s, the government systematically encouraged import substitution industrial growth, and nurtured an industrial elite dependent on state patronage. This was done by means of high protection rates to domestic manufacturers, cheap credit, and direct as well as indirect import controls on competing imports. (Kemal 1999, Soligo and Stern 1965, Rahim 2001) The wide range of protection measures and concessions provided by the government during the 1960s enabled the industrial elite to make large rupee profits without the market pressures to diversify into high value added industries or to achieve international competitiveness. These tendencies persisted in varying degrees for the next four decades5 . Yet they were at an economic cost that became a growing burden on the economy, particularly on the poor6 . The rapid increase in budget deficits associated with subsidies and protection measures obliged the government to undertake fiscal measures that constituted a double squeeze on the poor: Development expenditure that had provided a cushion to the poor against growing income inequality was drastically reduced from 7.4 per cent of GDP in 294 Akmal Hussain and Savail Hussain the 1970s to 3.5 per cent in the late 1990s. (Hussain 2003: 46). At the same time, the attempt to increase revenues through indirect taxation resulted in a further constriction of the real incomes of the poor (RTPA 1997). The Ayub period illustrates the historically rooted tendency of the government to seek political support amongst nascent elites through state patronage, even though the financial cost of such patronage added to the relative burden of the poor. In the subsequent Z. A. Bhutto period (1971–77), one of the most important initiatives of the government was the nationalization in 1972 of 43 large industrial units in the capital; the intermediate goods sector followed later by the nationalization of smaller industries such as cooking oil, flour milling, cotton ginning and rice husking mills. While the first set of nationalizations impacted the ‘monopoly capitalists’, the second set of nationalizations in 1976 hit the medium and small sized entrepreneurs. Therefore nationalization in this regime cannot be seen as state intervention for greater equity as was officially claimed (Noman 1988: 80). Rather, the rapid increase in the size of the public sector served to widen the resource base of the regime for the practice of traditional form of power through state patronage (Hussain 2004). General Zia ul Haq who overthrew the Bhutto regime in a coup d’état aimed to acquire a political constituency amongst the conservative religious strata of the lower middle class. This was part of his attempt to restructure state and society into a theocracy. The institutional foundation of what later came to be known as ‘Islamic fundamentalism’ was laid when government funds were provided for establishing mosque schools (madrassas) in small towns and rural areas, which led to the rapid growth of militant religious organizations7 . During the Zia period (1977–88), there was a strategic shift from the ‘socialist’ policies of nationalization and the large public sector in the Bhutto period to denationalization and a greater role assigned to the private sector in the growth process. In this context, the Zia regime offered a number of incentives to the private sector such as low interest credit, duty-free imports of selected capital goods, tax holidays and accelerated depreciation allowances. These incentives combined with high aggregate demand associated with consumption and housing expenditures from Middle East remittances, induced an increase in private sector investment and accelerated GDP growth during the period8 . The consequent increase in fiscal Poverty, Power and Local Government 295 space was used to win the political support of various echelons of the religious theocracy by using state funds to support madrassas. In the decade of the 1990s, financial resources from the nationalized banking sector were systematically used for political purposes. This was done by granting loans from the nationalized banking sector as political favours to individuals, many of whom defaulted on the loans9 . At the same time, state resources were used to grant contracts and licenses to enrich political allies. According to an estimate by Burki and Pasha (1999), the cost of such corruption to the banking sector alone was 10 to 15 per cent of the 1996–97 GDP during the period 1993–96 alone. The overall cost to the country of corruption at the highest level of government in the early 1990s has been estimated at 20 per cent to 25 per cent of the 1996–97 GDP, or approximately USD 15 billion (Burki 1999: 178). Occurring at a time when GDP growth had already begun to fall below its historical trend rate, widespread governmental corruption may have been a significant factor in intensifying the slowdown in investment, increasing the economic burden on the poor and perpetuating the inadequacy of basic services during this period (Hussain 2004). Corruption during the 1990s may have not only slowed down investment and growth but also increased inequality and the economic burden on the lower income groups. The problem of low revenues was accentuated by the massive leakage in the tax collection system due to corruption. This leakage amounted to 3 per cent of GDP, which was about twice the level 10 years earlier (Burki 1998). Since the government was unable to plug the leakage in the tax collection system or reduce non-development expenditure, it had to resort to increased indirect taxation to deal with the emerging fiscal crisis. Evidence on the incidence of taxation during the 1980s and early 1990s shows that the tax burden as a percentage of income was highest at 6.8 per cent for the lowest income group and lowest at minus 4.3 percent for the highest income group (RTPA 1997). Thus the indirect taxation, partly necessitated by governmental corruption served to accentuate income inequality and poverty. Thus the patron–client model of governance established during the British Raj continued in the post independence period in a variety of new forms. While in the British period, state resources were used as patronage to build political support for the Raj, state resources were used by individual rulers in the post independence period to 296 Akmal Hussain and Savail Hussain build ‘personalized’ domains of power. The process of constituting power by individual members of the ruling elite in the post independence period was integrally linked with the emergence of an economic structure characterized by endemic poverty. In the next section, we will discuss the counter posed forms of power manifested by the elite on the one hand and the poor on the other. POVERTY AND THE DIALECTIC OF POWER The ruling elite with rare exceptions practice a form of power that is counter posed to that of the poor (Hussain 2002). The power of the elite is constituted within the structure of patron–client relationships. At an economic level, it involves tying the poor individually into a chronic dependence on the patronage of the elite, operating within their individualized domains of power.10 At a psychological and social level, elite power involves creating a sense of powerlessness in their subjects: Internal relationships of fraternal loyalty and support within the community are ruptured, and the individual isolated and made dependent on the economic and social support emanating out of elite power. The exercise of this form of power involves constriction of the space for autonomous initiatives by the poor. Therefore, the power of the elite is predicated on the loss of freedom of the poor. By contrast the poor communities in Pakistan are imbued with a folk tradition where the process of actualizing the self is experienced through progressive integration with the community.11 Thus, empowerment of the poor involves a reintegration with their community. In contrast to the power nexus of the elite, when the poor are empowered, the isolation of the individual is replaced by integration with the community. This relatedness with the other and with the inner self creates a sense of freedom and opens the space for autonomous initiatives by the poor. Integral to this sense of freedom is the ability through community action to acquire better access to input and output markets, credit, training and government institutions for security and justice. Empowerment of the poor signifies relatedness, and acquiring the confidence and material basis for taking autonomous initiatives for development. Poverty, Power and Local Government 297 In the context of this dialectic, empowering the poor means breaking out of the nexus of elite power through a transformation of the economic, social and psychological conditions of the poor. One of the ways in which this can be done that has been demonstrated in a number of cases in South Asia, particularly Pakistan, is through Participatory Development.12 A brief description of this approach can be given as follows: Participatory Development is a process that involves the participation of the poor at the village level to build their human, natural and economic resource base for breaking out of the poverty nexus (Hussain 1994: 26–28). It specifically aims at achieving a localized capital accumulation process based on the progressive development of group identity, skills development and local resource generation. The essential features of Participatory Development include social mobilization, training, and participation within community organizations for development projects, small irrigation schemes, hygienic drinking water, health care and education. Social mobilization or group identity development proceeds through the initiation of a series of dialogues with rural communities. These dialogues culminate in the formation of community organizations of the poor, which undertake a series of projects for income generation and infrastructure development. Acquisition of new skills and active participation within their community organizations (in both the planning and implementation of projects) allows the poor to exert a new power over the economic and social forces that fashion their lives. DECENTRALIZATION AND EMPOWERMENT: THE LESSONS FROM SOUTH ASIA In the preceding part we analyzed the differing forms of power practised in Pakistan by the elite and the poor respectively. In this part we will examine the issue of empowerment in the context of decentralization reforms being undertaken in South Asian countries. In the context of empowerment through decentralization, four major lessons emerge from recent research in South Asia (Hussain 1994: 26–28):298 Akmal Hussain and Savail Hussain First, formal decentralization of power in itself does not necessarily help the poor as pointed out by Upadhyay (2004) in the context of the Nepal case study. Empowerment of the poor, he argues, requires that formal decentralization be accompanied by a rigorous process of social mobilization. This involves raising consciousness and building organizations of the poor. It is only such a process that will enable the poor to acquire countervailing power. Without this dimension of countervailing power, decentralization will merely result in the appropriation of the ‘fruits of decentralization (by the elite) for their own narrow benefit’ (Maqsood Ali 2004). The second lesson emerging from the case studies is that if decentralization is to enable empowerment of the poor, it must be holistic, that is, incorporate political power, emergence of social consciousness and administrative and fiscal devolution. At the same time, it must reach down to the grassroots level through various intermediate levels with institutionalized participation of the poor in governance at every level. Third, the political dimension of decentralization must be inclusive and capable of absorbing what Upadhyay (Upadhyay 2004) calls, ‘diverse ethnic and other identity groups as equal partners occupying spaces in the polity’(Hussain 2003). He argues that centralized polity excludes such identities that may be a factor in ethnic strife and social polarization. While the poor, once organized, are able to generate new resources at the local level, yet, as participatory development is scaled up, internally generated resources may be insufficient. Therefore externally generated resources become necessary, but these have to be carefully applied through a sensitive support system that strengthens rather than weakens the autonomy of the organizations of the poor. Such a support system could be provided by a combination of apex NGOs, state institutions, banks and local governments.13 On the basis of the lessons drawn from the recent case studies of decentralization reforms in South Asia, the following proposition can be made: For the poor to be empowered at the local level, if the decentralization currently being undertaken by a number of South Asian countries is to be effective, it cannot simply aim at decentralization of ‘administrative’ functions within existing governance structures. Rather, decentralization has to create the space within which an institutionalized relationship can begin between autonomous organizations of the poor and various tiers of local government. Poverty, Power and Local Government 299 POWER DYNAMICS OF LOCAL GOVERNMENT IN PAKISTAN: THE ISSUE OF EFFICIENCY The idea of decentralizing governance is drawn from social science theory stretching back to the age of enlightenment. In recent times, economists such as Oates (1972) and Teibot (1956) have propounded the welfare gains of decentralization. The argument is based on a simple proposition: The allocation of public resources at the local level is more likely to conform to public welfare priorities and the delivery of basic services is likely to be more efficient in a situation where these administrative functions are being performed by elected government officials, close to and in full view of the electorate. Thus, ‘proximity’ to the electorate and ‘accountability’ to them impel the local government officials to seek public good. By contrast, unelected bureaucrats in a centralized administrative system are disciplined through the less effective device of service rules.14 The theory of course, looks at centralized and decentralized governance as alternative options. It does not take account of the transition process of moving from one to the other. More importantly, it ignores the issues of ‘power’ involved in the transition to effective local government. Pakistan’s case constitutes a vivid illustration of the dynamics of power in the transition process. In an important study on decentralization in Pakistan, Cheema et al. (2003), make two noteworthy propositions which are relevant for our discussion: (i) In Pakistan, historically, decentralization measures in terms of local government have always been undertaken by military governments, which, while centralizing political power in their own hands, seek political legitimacy through local government. (ii) While such ‘non-representative governments’ have undertaken reforms for ‘representative’ local governments, they have at the same time sought to establish control over local governments through the bureaucracy. Information gleaned from our interviews with local government officials at various tiers in Lahore and Multan respectively, suggests that a contention for power is taking place between elected district government officials and the provincial bureaucracy. This contention threatens to paralyze the effective functioning of local government even for the limited objective of providing more efficient service delivery. This contention flows directly from the history 300 Akmal Hussain and Savail Hussain of the patron–client model of governance in Pakistan. The latest ‘Devolution of Power’ program threatens the terrain of resource gratification and discretionary appointments and transfers on the basis of which the bureaucracy and the elected politicians in provincial governments had constituted their domains of pelf and privilege. Therefore a contention for power has ensued between the provincial bureaucracy and the local governments. This contention was intensified with the recent establishment of elected provincial governments, in which elected members of parliament at the provincial level contend with local governments for control over development funds and discretionary appointments of officials. The provincial bureaucracy has been able to appropriate the authority to appoint key officials known as Executive District Officers (EDO). Yet it is through these officials that the elected local governments are supposed to administer the allocation of development funds and provision of basic services such as health, education, sanitation and drinking water. Moreover, all officials in various public service departments in the district administration from grade 11 to 18, are also appointed by the provincial bureaucracy. Thus, while the elected local government officials have responsibility, they do not have authority. Their ability to improve the delivery efficiency of public services is severely constrained by the fact that they can neither transfer, nor appoint most of the officials who operate these services. To make matters worse, the resources made available to local governments and the professional expertise at their command are so inadequate, that they are unable to take even a minimal initiative to fulfil their election mandate of widening the coverage and quality of basic services15. Elected local government officials (known as Nazims) are reduced to ‘requesting’ the provincial bureaucracy to fill vacant posts in various schools and health care facilities or to transfer employees who fail to perform their duties. The resultant delays, the lack of control over EDOs, the severe shortage of resources and expertise combine to severely constrain the effective functioning of local governments. As a consequence of this contention for power, efficiency in the provision of public services far from increasing, may in fact have been reduced. According to one Tehsil Nazim in Multan district, the dominance of the provincial administration over the functioning of district governments has created a hybrid creature which is preventing efficient service delivery and limiting the effectiveness of elected members of the local government. Poverty, Power and Local Government 301 In the future, local governments in Pakistan can take one of three routes: (i) The district level governments may be rendered so dysfunctional that Nazims may begin to resign and in the subsequent elections, genuinely popular local figures may lose interest in local government altogether. Such a process could ultimately result in the failure of the ‘Devolution of Power Programme’. (ii) The local government system as it presently exists may continue to function at such a low level of efficiency that the efficiency gains conceived in the programme may become low or even negative. (iii) The current situation where Local Government (LG) elected officials have responsibility without appropriate authority and where they are starved of financial resources may be changed. In this case, local government officials may be granted authority over appointments, promotions and transfers of all personnel in the district administration. At the same time, adequate technical and financial resources could be made available to elected LG officials. In such a case, the power of the provincial bureaucracy to establish personalized patron–client factions and the tendency to appropriate economic rent could be transformed into efficiency gains associated with effective decentralization. The devolution of power program in Pakistan stands at crossroads. There is a dialectic between two forms of power at the local level— one that is derived from building a domain of dependency through the arbitrary use of state resources. Even as this form of power is at play, an alternative form of power could emerge over time—a form of power that is based on winning and maintaining public support by elected government officials. Such public support would be achieved through the effective functioning of local governments for the provision of basic services, poverty reduction and economic growth. In the next section, we attempt to explore this dialectic in terms of efficiency and power constraints. V. RENT-SEEKING, EFFICIENCY AND CONSTRAINTS TO POWER: LOCAL GOVERNMENT VS. PROVINCIAL BUREAUCRACY In this section, we build on the ‘grabbing hand models of government’ framework developed by Shleifer and Vishny (1998) and Shleifer and Blanchard (2000). 302 Akmal Hussain and Savail Hussain We use it for examining the dynamics of political and economic power in the context of recent local government reforms in Pakistan and their implications for social welfare. Underlying the grabbing hand analysis is the idea that politicians do not necessarily maximize social welfare and pursue their own selfish objectives. These can be completely different and often opposed to social welfare. Selfish objectives could include maintaining political and economic power (by staying in office), and personal and particular factional as opposed to social enrichment in general. A democratically elected government for example, although constrained by the need for re-election, can often pursue private gain at a high social cost. This could include establishing patron–client relationships with members of civil society, where resource transfers (say subsidies) flow from members of government to certain politically influential members of civil society, in return for political support to prop up the bureaucrats’/politicians’ power base. Such behaviour by both governmental and civil society members can be detrimental for productive and innovative activity and hence economic growth. The tendency to seek economic rents and political advantage using state instruments is costly in terms of social welfare and growth foregone, in two main ways. First, rent-seeking ‘mechanisms’ can be subject to increasing returns, which means that very high levels of rentseeking may be self-sustaining. For example, there can be a fixed cost to setting up a rent-seeking system, such as laws facilitating corruption (say lobbying cost for bureaucrats for pushing through a legal requirement of a particular regulatory regime). Once such government regulations are in place however, bureaucrats can cheaply impose bribes for enabling the entrepreneurs or private sector entities adversely affected by such regulations to avoid them. Similarly, rent seekers have strength in numbers. If a few people steal, they are likely to be caught; if many do, the probability that any one of them is caught is much lower and hence the return to stealing is higher. (Shleifer and Vishny 1998). Second, seeking political advantage through appointments, transfers, selection of public sector development projects that win the bureaucrat political support for the furtherance of his career can also set up allocation distortions in the economy. This is because the most talented people may be blocked from progressing up the hierarchy of the bureaucracy at the expense of the most influential people, thereby distorting the former’s effort incentives. Poverty, Power and Local Government 303 The social costs of such distortions in terms of lost output can be significant. Cross country evidence finds a robust negative association between investment and corruption (a proxy for rentseeking) across a broad cross section of countries (Mauro 1995). We examine two tiers of government, the provincial bureaucracy and local government. The un-elected provincial bureaucracy has traditionally sought to establish a domain of power within which it can construct and maintain patron–client relationships to acquire political advantage (such as furtherance of an individual’s bureaucratic career) on the one hand and economic rents (such as bribes) on the other. Individual members of the bureaucracy have sought to build factions of clients using various instruments. These include grants of permits, contracts, and exemptions associated within a regulatory economic regime; appointment and transfer of individuals in various tiers of government departments and public sector industries; utilization of government financial resources in terms of the selection and implementation of various development schemes and government administrative expenditures. Political alliance building through granting favours to individual members of civil society allows the provincial bureaucrats to build political support in civil society. These political alliances wielded by bureaucrats can prove valuable when made available to elected cabinet ministers and political leaders, and hence can be used by bureaucrats as a quid pro quo for career advancement. Similarly, bureaucratic control over appointments and transfers of personnel, regulatory mechanisms such as permits and involvement in development projects can provide bureaucrats rents in the form of bribes and career promotions unrelated with merit. Elected local government officials who have key administrative roles in the administration seek primarily to establish and maintain power based on broad based electoral support. This objective creates the imperative to improve the quality and coverage of development projects and public services within their constituencies. There would still be a tendency to build patron –client relationships to seek both private or factional gain as well as economic rent. However, this tendency would be restrained by the imperative for more efficient service delivery in a situation where proximity to their constituencies makes their actions both transparent and accountable through the electoral process. Proximity to their constituencies and the need for re-election thus constrains the ‘grabbing hand’ of LG officials 304 Akmal Hussain and Savail Hussain as they attempt to seek rents through misappropriation of state resources and/or bribes from the private sector. These additional features of accountability and the need for re-election begetting responsibility can set the LG administration apart from the provincial bureaucracy vis-à-vis political alliance building and rent-seeking. However, the extent of LG pursuit of power through relatively efficient utilization of public resources (for improved coverage and quality of various services and development projects) will depend on the scale of financial and technical resources it can draw upon, together with the authority over allocation of resources and degree of autonomy over management decisions available to them. SECTION VI: CONCLUSION In this paper, we have shown how since the British Raj and during the various regimes in the post independence period, economic policy was designed to enable the elite to use state resources for building a domain of dependency amongst various social strata. Over the last four decades, the patron–client model of governance has led to the emergence of an economic structure that restricts poverty reduction for given GDP growth rates and also constrains the growth potential of the economy from being realized. One of the approaches to poverty alleviation that has been demonstrated in South Asia is empowerment of the poor at the local level on the basis of participatory development. This is a process that involves the participation of the poor at the village level to build their human, natural and economic resource base for breaking out of the poverty nexus. The question that arises is, can such empowerment of the poor emerge out of the decentralization reforms currently underway in Pakistan? The lessons of decentralization in South Asian countries suggest that even within the constraints imposed by elite dominance at the national level, it must go beyond mere decentralization of administrative functions for the ongoing decentralization to work for the poor. Rather, decentralization has to create the space within which an institutionalized relationship can emerge between autonomous organizations of the poor and various tiers of local government. Poverty, Power and Local Government 305 The functioning of existing local government in Pakistan is characterized by a contention for power between elected district government officials and the provincial bureaucracy. This contention threatens to paralyze the effective functioning of local government even for the limited objective of providing more efficient service delivery to the poor. Our analysis of the differing incentives and constraints operating on the provincial bureaucracy and local government shows that the latter are more likely to pursue pro-growth and pro-poor policies. At the same time, elected local governments face a set of incentives that induce them to be relatively less corrupt and seek relatively greater social welfare. This is because elected governments are impelled by the imperative of electability and proximity to their electorate. In terms of incentives and constraints, these are more stringent than the service rules operating on the provincial bureaucracy. However, the necessary condition for the results indicated above is the availability of a set of complementary inputs—adequate authority and resources to the local governments. These could be provided within an institutionalized relationship between organizations of the poor and local government. Such a local governance structure could enable the poor to participate in the process of resource allocation and resource use at the local level, even though there is very limited space for empowerment at the national level where the power structure is still dominated by traditional elites practicing power within the patron–client model. APPENDIX: COMPARING THE EFFICIENCY OF LOCAL GOVERNMENT WITH THE PROVINCIAL BUREAUCRACY16 Model [A]: Growth and Effi ciency17 In this sub section, we model the relative efficiency of local government compared to the provincial bureaucracy in Pakistan in terms of service delivery and pro-growth policy objectives, using the framework developed by Shleifer and Blanchard (2000). 306 Akmal Hussain and Savail Hussain Think of government as having two levels, provincial bureaucracy and local government (LG). The LG is elected while members of the provincial bureaucracy are appointed and disciplined according to service rules. Each tier of government has two possible policy options. They can foster growth by limiting the transfer of state resources to local elites and other interest groups and allow the private sector to develop and focus on efficient service delivery. Alternatively, they can be anti-growth by transferring state resources to old, established elites, hence deterring the private sector as the initial rent holders are protected from competition, together with a poor service delivery record. Why might a LG choose the second policy option? There are two main hypotheses, the first—call it ‘capture’—is that local governments can be captured by the initial rent holders, primarily the old, local landed elite, local state structures (for example, revenue and police officials), traders or ‘firms’. In this view, local governments work to generate transfer of state resources to these entities and protect them from competition by new entities while efficient service delivery is ignored, in return for bribes and kick backs. The second view, call it ‘competition for rents’, is that the behaviour of local governments is the unintended result of administrative disorganization. Too many agencies, all uncoordinated, try to extract rents from private firms18, making it unprofitable to create or run a private business legally, thus impeding localized capital accumulation by the poor. Here service delivery as well as economic growth suffers. We define the key parameters of the model as follows: Let ‘b’ be private benefits to the LG of ‘killing growth’. Under the capture interpretation, b may reflect the transfers back from existing elites to LG in the form of bribes, cash or in kind payments. Under the competition for rents interpretation, b may reflect the cost to a LG of reducing individual bribe taking by local officials, for example, through enforcing costly reform measures. Let ‘y’ be the additional output under growth. With appropriate normalization, let y also stand for additional revenues available to the provincial bureaucracy and LGs under growth. Suppose that a fraction ‘a’ of additional revenues from growth go to LG. Thus additional revenues from growth to LGs are simply: ‘ay’. Assume further that a proportion ϕ of additional revenues from growth are misappropriated by government officials in the form of kick backs from discretionary spending. Thus for LG officials, the returns from growth are ϕ (ay). All benefits to government officials from pursuing pro-growth or antigrowth policies depend on them holding office. In addition to benefits associated with the type of policy option chosen, LG government officials by remaining in office, also derive benefit from the sense of respect that is associated with elected office. Let us call this measure of benefit from holding elected office kLG= T1/2, where T is number Poverty, Power and Local Government 307 of terms in office. This particular functional form for kLG, specifically k´LG > 0 and k´´LG < 0, captures the notion that the marginal returns to re-election are positive but diminishing. The justification for including kLG in the model is that each LG official values remaining in office through the electoral process. The bureaucratic official also derives pleasure from holding office. However, his stay in office is determined not by the voters but by his superiors in the bureaucratic hierarchy. We can take the benefit from holding a bureaucratic office as kB = X, where X is a constant. Since there is no conception of election to office, the pleasure of acquiring public office through the electoral process is absent for bureaucrats. Let py LG be the probability that a LG remains in office either by finishing his term or being re-elected, given that he follows pro-growth, efficient service delivery policies. Suppose further that px LG is the probability that a LG remains in power if it pursues anti-growth policies along with poor service delivery. Similarly, suppose py B is the probability that a provincial bureaucracy remains in office if it pursues pro-growth, efficient service delivery policies and let px B be the probability that it remains in power if it follows anti-growth policies together with poor service delivery. Let PLG = py LG /px LG and PB = py B /px B. Under these assumptions, an LG in a given term chooses growth if: py LG [ϕ (ay) + kLG] > px LG (b + kLG ) or alternatively PLG [ϕ(ay) + kLG] > b + kLG , (1) Similarly, during a particular term and given a, y, and b, a provincial bureaucracy will choose growth if: py B [ϕ (ay) + kB] > px B (b + kB ) or alternatively PB [ϕ (ay) + kB] > b + kB , (2) Which tier of government is more likely to pursue pro-growth, efficient service delivery policies? To answer this question, we have to delve into the nature of the set of rules that govern an official’s stay in office, be it LG or provincial bureaucracy. Since LG officials are elected and electoral success can safely be assumed to depend upon delivering pro-growth, efficient service delivery policies, PLG is likely to be large both because of a high py LG and a low px LG. This hypothesis accords well with intuition: An LG is more likely to be re-elected if it adheres to pro-growth policies and less likely to remain in office if it follows anti-growth policies. The fact is that the proximity of LGs to their electorate adds to transparency, and hence greater accountability of government adds weight to our argument for a large value of PLG.308 Akmal Hussain and Savail Hussain The tenure in office of a particular official in the provincial bureaucracy is governed by a different set of rules: not the electoral process driven by performance, but rather by service rules. Service rules are generally accepted to be less stringent than the electoral process when it comes to rewarding or penalizing governmental performance. Furthermore, the bureaucracy has a comparative disadvantage compared to LG when it comes to proximity with the people who are affected by their policies. The boost that proximity gives to accountability (via greater transparency) of government is thus weaker for provincial bureaucracy as opposed to LG. We can then take PB to be relatively lower than PLG. Specifically, py B can be taken to be close to 0, since what determines a bureaucrat’s stay in office is not how well he delivers in devising and implementing pro-growth policies and efficient service delivery, but rather how effectively he can lobby the upper echelons of the bureaucratic hierarchy who control his appointment within and movement up the bureaucratic order. This also implies that poor performance in terms of development is less likely to be penalized within a bureaucratic order compared to LG who can be voted out in the next election cycle. Thus px B is likely to be higher than px LG. Our analysis then leads us to the following inequality: PLG = py LG /px LG > PB = py B /px B , (3) Applying (3) to (2) for given a, b, kLG and kB and taking py B = 0 we find that: PB [ϕ (ay) + X ] < b + X THE ABOVE INEQUALITY IMPLIES THAT THE PROVINCIAL BUREAUCRACY WILL UNAMBIGUOUSLY PURSUE ANTI-GROWTH POLICIES GIVEN B > 0. Given that PLG is > 0, an LG is thus more likely to pursue pro-growth, efficient service delivery policies than a provincial bureaucracy. Model [B]: Inter-Temporal Rent-Seeking19 In this sub section, we use and extend the analytical framework developed by Shleifer and Vishny (1993) by using a multi period setting to demonstrate how rent-seeking under LG is likely to be lower than under the provincial bureaucracy in Pakistan. Poverty, Power and Local Government 309 Suppose there are two government produced goods such as building permits x1 and x2. The former can only be sold in period one and the latter only in period two. We assume there is demand for these goods is D(pi ) from private agents, where i, is an index for time period. Assume further that the price elasticity of demand for both permits is less than 1. We assume that these goods are sold for the local government by an official who has monopoly over supply and who thus has the opportunity to restrict quantity of the good that is sold in each period in order to raise its price and hence earn bribe revenues or rent. Let the price of the building permit x1 be p1 and let the price of the building permit x2 be p2 . We assume that the official has to turn in the official government price to the state treasury after each sale. The official government prices for the building permits in periods one and two can then be viewed as the official’s marginal cost of supplying the permits. Let the official government price of the building permits in periods one and two be MC1 and MC2 respectively. For the sake of clarity, imagine the governmental official in-charge of granting the permits as an agency that produces and sells building permits in two periods at prices p1 and p2 at marginal costs of MC1 and MC2 . The official remains in office for a maximum of two time periods t1 and t2 , and can sell one permit per period. Whether his tenure in office extends to t2 depends on his re-election to office. Re-election in turn is determined by how far the official can limit corruption in his administration, and how well he delivers on service delivery and whether he pursues pro-growth policies. In order to maximize rents from both types of permits over the two time periods, the official is constrained from maximizing bribe revenues in the first period from permit x1 , for doing so means he won’t be re-elected and thus will not extend his tenure in office into the second period. For analytical purposes then, we can treat x1 and x2 as complimentary goods from the official’s perspective, for how x1 is priced will determine the demand for both x1 and x2 as far as the official is concerned. Specifically, if the bribe price of the permits in period one is set too high, then the official will be voted out of office and will not earn any returns on the permits in period two, that is, as far as the official is concerned, demand for x2 will be zero. Formally then if the official wants to maximize joint revenues from x1 and x2 , he will set p1 such that: [ ] ,( ) MR dx MR dx r MC 2 2 1 1 1 1 + 1 + = (1) We then take dx dx 2 1 > 0 given the complementarity between x1 andx2 and r is the interest rate.310 Akmal Hussain and Savail Hussain (1) implies that at the optimum MR1 < MC1 . The monopolist agency keeps the per unit bribe (P1 – MC1 ) on the permit in period one down to expand the demand for the complimentary permit in period two. Given an N period time horizon, the monopolist agency will keep the bribe down on building permits for N-1 periods. Suppose alternatively that the building permits are allocated by a bureaucrat. Suppose, alternatively, that the building permits are allocated by a bureaucrat. A bureaucrat’s stay in office as we discussed earlier is not determined through the electoral process but rather by a set of service rules and members of the bureaucratic hierarchy sitting above the particular bureaucrat in question. Service rules are less rigorous when it comes to providing appropriate incentives and disincentives for those in the public domain to ensure improved service delivery and pursue pro-growth policies. Progress up the bureaucratic hierarchy also involves lobbying senior members and again performance within the public domain does not have to be the driving force behind a bureaucrat’s promotion. Members of the bureaucracy are also at a comparative disadvantage compared to LG members when it comes to proximity with the beneficiaries of their policies. Accountability is thus lower for bureaucrats due to reasons of transparency than LG officials. Furthermore, members of the bureaucracy compared with officials of LG have a comparative disadvantage when it comes to proximity with the beneficiaries of their policies. Accountability is thus lower for bureaucrats due to reasons of transparency than LG officials. Given these arguments, (1) would have to be modified if the bureaucracy rather than the LG was in charge of issuing permits x1 and x2 . Specifically, the absence of the electability constraint and the lower public accountability of the bureaucrat (supplying permits on behalf of the bureaucracy) implies that he can maximize bribe revenues from each permit in each period without fear of dismissal, for dismissal is unlikely to be related to performance but depends on the bureaucrat’s effectiveness in lobbying the upper echelons of the bureaucracy and his ability to work around service rules that regulate his tenure in office. In fact, given that lobbying costs are positive and likely to be increasing for higher posts in the bureaucratic hierarchy, there will be strong incentives for bureaucrats seeking promotion to earn as much as they can from a particular office in a bid to lobby up the hierarchy. We then take dx dx 2 1 = 0, when dealing with a bureaucrat, for as far as he is concerned, there is no complimentary relationship between x1 and x2 . The Bureaucrat maximizes his objective function when MR1 = MC1 Hence the per unit bribe on the permits in period one is going to be higher and output lower than at the joint monopolist optimum. Given N periods, the above optimality condition applies to permits in all N periods.Poverty, Power and Local Government 311 Our model of inter-temporal rent-seeking suggests that the rents sought by LG officials are lower than those appropriated by bureaucrats. The main reasons why rent seeking will be higher under the provincial bureaucracy than LG are the absence of the electability constraint on bureaucrats together with lower public accountability stemming from lesser transparency. NOTES 1. For data and analysis on the processes of poverty within the structure of elite power at the local level, see Hussain et al. (2003). 2. For a detailed discussion with supportive evidence, see Ali 1988. 3. Unlike most other parts of pre-British India, in Punjab, the upper echelons of the social hierarchy were extensively displaced at the end of the Mughal period. The growing momentum of peasant uprising in the 18th century led by land holding segments of the peasantry, culminated in these peasant war bands asserting autonomous control over land and political authority. The upper peasant rebel leadership emerged as a new class of superior land holders, who with the onset of the British Raj, were later acknowledged as such by the British. Evidence of the displacement of the older Mughal period elite comes from British documentation such as the District Gazetteers and Griffin and Massy’s tome, Chiefs and Families of Note in the Punjab. Very few families identified at the district level as of elite status, had such antecedents prior to the late 18th century. See Ali 2002. 4. For a more detailed discussion, and evidence, see Ali 1980., Ali (1987), Ali (1988). 5. That protection to the large-scale manufacturing sector persisted for the next three decades is indicated by the fact that even in 1990–91 by which time the rate of effective protection had been considerably reduced, the increase in the share of manufacturing attributable to protection amounted to 5 per cent of the GDP. Similarly the failure of the industrial elite to diversify exports into high value added non-traditional industries is indicated by the fact that the textiles and related goods sector constituted 5 per cent of commodity exports in the period 1960–70, and continued to remain as high as 50 percent three decades later in the period 1988–99. See: Hussain 2004. 6. During the 1960s, while a highly monopolistic elite was amassing wealth, the poor in Pakistan were experiencing an absolute decline in living standards. This is indicated by the fact that the per capita consumption of food grain of the poorest 60 per cent of Pakistan’s urban population declined from an index of 100 in 1963–64 to 96.1 in 1969–70. The decline was even greater over the same period in the case of the poorest 60 per cent of rural population. In their case, per capita consumption of food grain declined from an index of 100 in 1963–64 to only 91 in 1969–70. (see Hamid 1974). There was an even larger decline in the real wages in the industry: In the decade and a half ending in 1967, real wages in the industry declined by 25 per cent. (See Griffin and Khan). According to one estimate, in 1971–72, poverty in the rural sector was so acute that 82 per cent of rural households could not afford to provide even 2,100 calories per day per family member. (See Naseem 1977).312 Akmal Hussain and Savail Hussain 7. This social process was catalyzed by the Afghan war. Zia sought political, economic and military support from the US by offering to play the role of a front line state in the Afghan guerrilla war against the occupying Soviet army. Accordingly, Pakistan obtained a package of US$ 3.2 billion in financial loans. Additional fiscal space was obtained by getting foreign debt rescheduled and increased private foreign capital inflows. These official and private capital inflows played an important role in stimulating economic growth in this period. They also helped establish a political constituency both within the institutions of the state and in the conservative urban petit bourgeoisie for a theocratic form of military dictatorship. 8. Private sector gross fixed investment increased from 7.1 per cent of the GDP in the Bhutto period to 9.2 per cent in the Zia period. At the same time, GDP growth accelerated from 4.9 per cent in the Bhutto period to 6.6 per cent in the Zia period (Hussain 2004: Table 3). 9. For example, S. J. Burki has shown that during the mid 1990s, large amounts of funds were siphoned off from public sector banks, insurance companies and investment institutions such as the National Investment Trust (NIT) and the Investment Corporation of Pakistan (ICP). The evidence was found in the nonperforming loans, which the state controlled financial institutions were forced to give to the friends of the regime, in most cases without collateral. During this period, the NIT and ICP were forced to lend to patently unviable projects, which were then quickly liquidated. The purpose of such lending apparently was not to initiate projects but to transfer state resources into private hands. The case of an oil refinery in Karachi and a cement plant in Chakwal have been quoted as examples of infeasible projects funded by the NIT on political grounds with both projects declaring bankruptcy (see Burki 1999: 175). 10. For a discussion based on survey data that shows how poor peasants tied to the landlord get less than market wage rates when working as day labourers on the landlord’s owner cultivated farms, see Hussain 1980: 363–81. For a more recent discussion based on the latest survey data, see Hussain et al. 2003: 62–63. 11. For an analysis of this phenomenon based on the Punjabi Sufi tradition, see Syed 1969. 12. (i) For a first hand field experience account of a Participatory Development initiative in ten districts of the Punjab, see Hussain 1998. (ii) For an articulation of the theory of Participatory Development and illustrative case studies of initiatives in the field in a number of South Asian countries, see Wignaraja et al. 1991. 13. For details see: Wignaraja and Sirivardana 2004. 14. For a review of the literature on the economic arguments for decentralization, see Tanzi 1995. 15. According to the Nazim of city district Lahore, the financial resource constraint was so severe that only rupees 70 million were available for new development, a sum which was not even enough for painting traffic markers on the roads of Lahore city. At the same time, almost all the local government officials interviewed reported an absence of adequate technical expertise within their administration and hence a severe constraint in designing development projects. (The project proposal form that is required to be submitted for funds allocation runs into about a hundred pages and requires specialized technical expertise to fill). Poverty, Power and Local Government 313 16. For a detailed examination of this point see Shleifer and Vishny 1993. 17. This section draws and builds on analytical frameworks developed by Shleifer and Blanchard (2000) and Shleifer and Vishny (1993) 18. This sub-section draws and builds on Shleifer and Blanchard (2000). 19. For a detailed examination of the link between coordination of corruption activities and social welfare, see Shleifer and Vishny. 1995. 20. This part of the sub-section adapts and builds on Shleifer and Vishny 1993. REFERENCES Alavi, H. 2001. ‘Parting of the Ways’, in Z. Mustefa (ed.), The South Asian Century. Karachi: Oxford University Press. Ali, Imran. 1980. ‘The Punjab Canal Colonies, 1885-1940’, Ph.D. thesis in Australian National University. Ali, Dr. Shaikh Maqsood. 2004. ‘The Bangladesh Case Study’, in P. Wignaraja and S. Sirivardana (eds), Pro Poor Growth and Governance in South Asia: Decentralization and Participatory Development. New Delhi and London:SAGE Publications. Ali, Imran. 1987. ‘Malign Growth? Agricultural Colonization and the Roots of Backwardness in the Punjab’, Past and Present, 114, pp. 10–132. Oxford. ———. 1988. The Punjab Under Imperialism, 1885–1947. Princeton, New Jersey:. ———. 2002. ‘Government Patronage and Rent Seeking Elites: A Longer Historical Perspective’, Background paper (Mimeo) for Pakistan National Human Development Report, UNDP, 2003. Burki, Shahid Javed. 2004. Pakistan: Fifty Years of Nationhood. Lahore: Vanguard. Cheema, Ali, Asim Ijaz Khwaja and Adnan Qadir . 2003. Unpublished. Decentralization in Pakistan: Context, Content and Causes, (Mimeo). Griffin, K. and A.R. Khan. 1972. Growth and Inequality in Pakistan. London: Macmillan. Hamid, N. 1974. ‘The Burden of Capitalist Growth, A Study of Real Wages in Pakistan’, Pakistan Economic and Social Review, 12(1): 47–70. Hussain, Akmal. 1980. ‘Impact of Agricultural Growth on Changes in the Agrarian Structure of Pakistan with Special Reference to the Punjab Province’, pp. 363–81. D. Phil Thesis, University of Sussex. ———. 1994. Poverty Alleviation in Pakistan. Lahore: Vanguard. ———. 1997. Overcoming Poverty. Report of the Task Force on Poverty Eradication, Islamabad. ———. 1998. Punjab Rural Support Programme, The First Four Months. Report submitted to the Board of Directors, PRSP, November. ———. 2002. Pro-Poor Growth, Participatory Development and Decentralization: Paradigms and Praxis, (Mimeo). Hussain, Akmal et.al. 2003. Poverty, Growth and Governance, Pakistan National Human Development Report, 2003, Chapter 3. Karachi: UNDP, Oxford University Press. Hussain, Akmal. 2004. ‘Institutions, Economic Structure and Poverty’, South Asia Economic Journal, 5(1): 73–74. Institute of Policy Studies, Colombo. India: SAGE Publications. 314 Akmal Hussain and Savail Hussain Kemal, A.R. 1999. ‘Patterns of Growth in Pakistan’s Industrial Sector’, in Shahrukh Rafi Khan (ed.). Fifty Years of Pakistan’s Economy, p. 158. Karachi: Oxford University Press. Mauro, P. 1995. ‘Corruption and Growth’, Quarterly Journal of Economics, 110(3): 681–713. Naseem, S. M. 1977. Rural Poverty and Landlessness in Asia. ILO Report, Geneva. Noman, Omar. 1988. The Political Economy of Pakistan, 1947–85, p. 28 London: Routledge Kegan and Paul. Oates, Wallace. 1972. Fiscal Federalism. New York: Harcourt Brace Jovanovich. Pasha, M. 1998. Colonial Political Economy: Recruitment and Underdevelopment in the Punjab. Karachi: Oxford University Press. Rahim, Sikander. 2001. ‘Myths of Economic Development’, Occasional paper no. 10 submitted in Lahore School of Economics, Lahore, February. Syed, Najam Hussain. 1969. Recurrent Patterns in Punjabi Poetry, Lahore: Majlis Shah Hussain. ———. 1986. Recurrent Patterns in Punjabi Poetry, Lahore: Punjabi Adabi Markaz. Shleifer, A. and O. Blanchard. 2000b. ‘Federalism with and without political centralization. China versus Russia’, NBER Working Paper No. w7616. Shleifer, A. and R. Vishny. 1993. ‘Corruption’, Quarterly Journal of Economics, 108(3): 599–617. ———. 1998. The Grabbing Hand. Cambridge, Massachusetts, USA:Harvard University Press. Soligo and J.J. Stern. 1965. ‘Tariff Protection, Imports Substitution and Investment Efficiency’, The Pakistan Development, pp. 249–70. Syed, Najam Hussain. 1986. Recurrent Patterns in Punjabi Poetry, Second Edition. Lahore: Punjab Adbi Markaz. Tanzi, Vito. 1995. ‘Fiscal Federalism and Decentralization: A review of some efficiency and macro-economic aspects’, Annual Conference on Development Economics, World Bank. Teibot, Charles. 1956. ‘A Pure Theory of Local Expenditures’, Journal of Political Economy, 64: 416–24. Upadhyay, S.K. 2004. ‘The Nepal Case Study’, in P. Wignaraja and S. Sirivardana (eds.), Pro Poor Growth and Governance in South Asia: Decentralization and Participatory Development. New Delhi and London: SAGE Publications. Wignaraja, Ponna, Akmal Hussain, Harsh Sethi, Ganeshan Wignaraja. 1991. Participatory Development: Learning from South Asia. Karachi: Oxford University Press and United Nations University. Wignaraja, Ponna and S. Sirivardana. (eds). 2004. Pro-Poor Growth and Governance in South Asia, Decentralization and Participatory Development. New Delhi and London: SAGE Publications.