Introduction

Public policy studies have emphasised different approaches to the study of policy processes. The term policy process means that public policies have a life of their own once the decisions on them are taken. This view emphasises that public policy making should be sensitive to the contingencies of socio-historical context as well as the institutional and organisational context in which policies are made. The view of public policy making as a policy process attempts to sensitise the policy makers and implementers to both the socio-historical complexity as well as to the institutional/organisational context. Both form the ‘policy environment’ in which policies are made and implemented; policy processes are inextricably contextualised in both these senses. The success, failure or partial successes of public policies depend on how much of the socio-historical and institutional/organisational context is considered at the time of formulation of public policies. Mooij provides a useful definition of viewing policies as policy processes. According to her:

‘The term ‘policy process’ is based on the notion that policies are formulated and implemented in particular social and historical contexts, and that these contexts matter—for which issues are put on the agenda, for the shape of policies and policy institutions, for budget allocations, for the implementation process and for outcome of the policies. The idea of ‘policy process’ stands opposed to that of ‘policy as prescription’ which assumes that policies are the result of a rational processes of problem identification by a benevolent agency (usually the state)’ (Mooij and de Vos 2003)

Two main approaches that figure prominently in the policy process literature particularly in the context of economic and governance reforms are the linear model of policy process and the interactive model of policy process (Turner and Hulme; 1997). Thomas and Grindle (1990) elaborate this view of linear and interactive or iterative models of policy processes in the third world reform context. According to them, the linear model of policy process views the operationalisation of a particular policy in linear fashion where an issue first figures on the policy agenda then decisions are taken by policy elites and the implementation follows. This model of policy process is also called ‘decisionistic’ policy process model, where decision is every thing and all else follows. The ‘interactive’ or ‘iterative model’ of policy process on the other hand problematises the linear model and emphasises the interactive, non-linear and political aspects of policy processes at all the stages of policy process: at the agenda stage; decision stage; and at the implementation stage, specifically the latter. Thomas and Grindle (1990) propose the ‘interactive model’ of policy process and recommend that the third world ‘reform’ policy processes are better understood through such a heuristic lens. They say:

‘The central element in the model is that a policy reform initiative may be altered or reversed at any stage in its life cycle by the pressures and reactions of those who oppose it. Unlike linear model, the interactive model views policy reform as a process, one in which interested parties can exert pressure for change at many points. Thomas and Grindle (1990, pp 1166–1167)

Basing their analysis on twelve case studies from across third world countries they specifically focus on implementation stage of policies and say that:

‘Our observations over many years as well as our research indicate that implementation is often the most crucial aspect of the policy process and that the outcomes of implementation efforts are highly variable ranging from successful to unsuccessful, but including also an almost limitless number of other potential outcomes’ (Thomas and Grindle 1990, pp 1164–1165)

Thomas and Grindle provide two alternative models of policy process in the first one, which they call ‘linear model’ the progression of the policy from policy issue to implementation phase is linear. It passes through the agenda phase, decision phase towards implementation phase without any interaction with social context and unknown outcomes.

Whereas in the second model, which they call the interactive model, the policy process is quite elaborate and inclusive of many factors. For example, a policy issue is taken up after due process they come on to the policy agenda following through decision and implementation phases. But the crucial point of the interactive model is policies when they go through the policy process in a political system, always have ‘multiple potential outcomes’. This is so because in the policy process the policy issues face the social and political contextual factors.

These approaches explaining policy processes are now applied to developing countries and policy processes particularly in the context of economic and governance reforms. It is argued that many specific policies that are envisaged by different state agencies now can be explained by these approaches in general and by the latter interactive/iterative (these words are used synonymously throughout the paper) model in particular. The policy process as an iterative process means that different actors in society and polity shape this policy process. We have seen in the above statement of Mooij that different actors shape the policy process. Here it would be appropriate to mention further about this aspect. Mooij for example notes after an exhaustive review literature on policy processes in the Indian context that:

‘[Policy processes literature, both theoretical and empirical] does not assume that policies are ‘natural phenomena’ or ‘automatic solutions’ resulting from particular social [or political] problems and it does not privilege the state as an actor fundamentally different from other social actors. The why, how and by whom questions are treated as empirical questions; it is only concrete empirical research that can generate these answers.’ (Mooij and de Vos 2003).

This paper argues that the interactive or iterative approach of policy process is applicable to decentralisation policy in India as well; particularly the interactive model of policy process is usefulFootnote 1 and argues that decentralisation in India should be seen as an iterative and interactive process. In the following, we look at how decentralisation process in India involves and is shaped by multiple actors as well as agencies in an iterative manner. We will then look at the nature and characteristics of the different actors that shape this policy process. At the end, we attempt to pull the different strands of the argument together.

Decentralisation as a policy process: the argument

There are only few studies from policy process prism of view regarding decentralisation. This article may well be first of its kind. There are studies on many other fields viewing the subject from the policy processes’ prism. For example, the paper by Keeley and Scoones (1999) ‘Understanding Environmental Policy Processes: A Review’ these authors deal broadly with the relationship between science, knowledge processes and environment policy; Jos Mooij’s unpublished paper 2002a, b) ‘Smart Governance: Politics in the Policy Process in Andhra Pradesh, India’; James Manor’s (1995) ‘The Political Sustainability of Economic Liberalization in India’; Myron Weiner’s (1999) ‘The Regionalization of Indian Politics and its Implications for Economic Reform’; Ashutosh Varshney’s paper (1999) ‘Mass Politics or Elite Politics? India’s Economic Reform in Comparative Perspective’, these deal with politics of reforms from the perspective of policy processes; G. Thimmaiah’s (1996) paper ‘The Political Economy of Populist Programmes’, Thimmaiah’s paper deals with how populism some times may result in distributive politics via unintended consequences of policy as manifestation of policy process. There is urgent need for research from the policy processes perspective into decentralisation policy in all three areas of rural decentralisation, urban decentralisation and the Panchayat Extension to Scheduled Areas (PESA) Act policy and implementation. This article is a modest invitation to decentralisation scholars to shed strictly an institutional perspective and venture into the viewing of decentralisation as basically a policy process whose effects often are more than the intended policy and also often different from the intended policy.

Decentralisation in Indian democracy, as is well known, has taken a new lease of life after the 73rd, 74th Constitutional Amendments and the PESA (Panchayat extension to scheduled areas) Act of 1996. The state governments were expected to pass conformity laws in accordance with the former. The linear logic would have it that the Parliamentary Legislation would be adopted by the states and implementation by the state level governments would follow. Local self-governments would then be implemented in rural, urban and scheduled areas. In contrast to this picture, the policy when adopted and implemented is rarely in conformity with the letter and spirit of the Constitution. The conformity laws are implemented according to the wishes of the governments at the state level. This is well documented (Jayal et al. 2007, Jha and Mathur 1999; Mathew 2000). But interestingly much policy process and policy politics take place even prior to the implementation of the legislations at the state level.

Soon after the laws to invest the local governments with the Constitutional status are passed, the Parliament also passed the legislative fiat on the Member of Parliament Local Area Development Scheme and most of the areas in which the Members of Parliament can intervene are the areas in which the local governments are expected to have powers over self-government. Therefore, the policy process initiated to constitute local government is over taken by the political compulsions to enable MPs their control over patronage linkages, clientele and vote banks (Mathew 2000; Jayal et al. 2007). Therefore, from the outset, the decentralisation policy in India is a non-linear, iterative process.

Secondly at the state level, while the conformity laws are adopted, the ‘Area Development Funds’ to members of legislature are given. More commonly, as is again well debated, the devolution of functions, funds and functionaries to local governments is at best halfhearted owing to the unwillingness of state level governments, including both politicians and bureaucracy. These are well noted in the literature (Joseph 2007; Mathew 2000; Mathur 1996; Jha and Mathur 1999; Jayal et al. 2007; Jain: n.d.). What is less noted is that there are multiplicity of actors and agencies shaping or influencing the policy process. The entire policy process is shaped by state as well as non-state actors, including civil society actors (prominently, NGOs).

At the state level, the first and foremost actors that influence both positively and negatively in accordance with circumstances are the bureaucracy. Often the Indian Administrative Service officers heading the line departments and the local level bureaucrats show either resistance, or ambivalence, at best ambiguous response to the laws of local self-government. The control over finances and functions and functionaries, if not opportunities to rent seeking, are some of the many reasons. Often simple—or perhaps complex—reluctance to give up power makes bureaucracy uncooperative to decentralisation process.

The less acknowledged agencies in the decentralisation policy are the political parties. Often in a number of states the local self-government elections are fought on political party lines. Political parties are deeply interested in winning the local body elections. This applies to parties across the political spectrum. The parties that win local government elections attempt to retain their control. And those that have lost constantly attempt to expand their party bases, cadres and following at the local levels. With local body elections, democracy surely reaches the grassroots level in urban, rural and tribal areas; the question, however, is whether this also serves the developmental needs of the local communities. Developmental outcomes through deepened democracy need not follow automatically. What, however, is vindicated is the fact that the decentralisation process becomes a complex policy process influenced by all these actors.

Attention to the role of political parties in decentralised democratic policy process is underpaid. Hasan writing in Indian context notes that ‘parties are seen, both by their members and by others, as agencies for forging links between citizens and policy makers. Their raison de tre is to create a substantive connection between the ruler and the ruled’ (Hasan 2002, pp 4–5). The same holds true for local body elections. Political parties in Indian context view local government elections almost plebiscitary, or as significantly indicative, of their political bases and attach considerable importance. If the ruling parties attempt to win over local citizenry through multiple developmental as well as populist schemes, the opposition parties start building their vote banks from the local level. The result often is a local community with fragmented political loyalties. The much idealised communitarian spirit of local self-government and cooperative local development of the communities is lost in the process. The political process even fragments the marginalised and weaker sections of local communities in both urban and rural areas. This happens even when the last tier of the local government, either an urban ward or a Grama Panchayat is officially declared to hold a party-less election. Even in a supposedly party-less election at the local level, political parties put up candidates for election and mobilise electorate on political party lines. This makes the local government as well as local development enormously politicised. If anything, in a face-to-face local community, political competition only becomes more intense than at higher levels of the polity. In an election to state assembly or national Parliament the candidates are distant, their victories and losses are felt only indirectly, whereas in a local body election the results and meanings of electoral battles concern the citizens directly and immediately. The decentralisation policy process therefore on one hand is positive in deepening democracy; on the other hand, the process robs local communities of their communitarian spirit. Whether or not to lament on the latter is a more normative question.

The other relatively new entrants into the complex policy process are the NGOs/civil society agenciesFootnote 2. These have been attempting to infuse the weaker sections of the local communities (Rai 2001) with confidence to take part in the democratic process via capacity building programmes, exposure visits, training camps on different themes of local government and various such handholding methods. These agencies have also been attempting to make local bureaucracy accountable to local representatives and help development funds reach the intended beneficiaries. But the NGO/civil society actors, even at their best, have serious problems of scale. The civil society actors cannot reach the scale at which the support to, and strengthening of, local government and local development are to be carried out. The continental scale of local democracy, and its depth in terms of reach, poses severe challenges to NGOs/civil society groups. In this context, some of the NGOs have been trying to enter into partnership with government agencies in the above-mentioned activities; and that opens discussion on a policy process within the larger policy process of decentralisation. The potential, therefore, of NGOs/civil society actors lies in strategic influence on policies, actors and agencies at higher levels of the political system in moulding the policy process in favour of women, weaker sections and the marginalised in the local contexts. Thus, the NGOs/civil society actors/activist groups/advocacy groups have emerged as strategically important players in the policy process of decentralisation. These can no more be ignored while reflecting on decentralisation as an iterative policy process. In addition to the above discussed political parties and NGOs, there have also been others such as micro-credit groups and others that have come to play their role both in local government and development. We consider them below.

Another aspect that reinforces the decentralisation process as an iterative policy process is the attempts of the state governments to sideline or bypass the local government through different ‘parallel bodies’ such as watershed committees, water users’ committees, women’s self-help groups, school education committees, vana samrakshana samitis (Joint Forest Management Groups) etc. These are often donor driven and financially better supported than the local government—and thereby attract more patronage, visibility and political clout. Real or bogus, these have been used by the state governments to render the democratic decentralisation process weak. The point of mentioning these as ‘damaging’ (Manor 2004) particularly in the context of this paper is that the state governments have, by the token of the above, rendered decentralisation a completely interactive policy process. Decentralisation policy once adopted is not adopted in a straight forward linear fashion, but goes on in a process of two steps forward and more than one step backward. That is what we intend to illustrate by presenting decentralisation as an interactive process, now-retreating, now-moving-forward, policy process. This is in addition to the always already present social context in India; the policy process takes place in the specific Indian social context, which adds its own twists and turns to the decentralisation policy.

Even in a positive situation of state showing all willingness to see the decentralisation policy through, even in a situation of less politicised rural communities, and absence of ‘parallel bodies’, even in such an ideal, favourable climate, the social complexity gives its own features to the decentralisation process. The realties of caste, class and gender ascribe to the decentralisation process their own marks. Local hierarchies prevail over local governance process both informally (Kripa AnanathPur 2004) and formally. The decentralisation process goes through the sieve of societal complexity resulting in a complex policy process. When we note the above, we should also note that the Indian polity at a macro level is increasingly witnessing the emergence of masses into politics (Yadav 2000; Hasan 2002; Frankel 2005). Hasan (2002), for example, notes that in Indian politics, ‘Expanding participation has placed the poor and the downtrodden groups in the caste, class and gender hierarchy at the centre of the political system’ (Hasan 2002, pp 22–23).

In such context, the local social groups are increasingly mobilised to enter the political process; this is happening more prominently at local level political process. This means both increasing political participation and political consciousness of the interests of different social groups also lead to frequent conflict. The story of decentralisation as a highly political policy process becomes clear.

One more policy processes aspect of decentralisation process in India is the Panchayat Extension to Scheduled Areas (PESA) Act of 1996. The PESA Act is basically meant to protect the interests of tribal communities of the country and the Act is flagrantly violated in its implementation across India. The tribal populations are most vulnerable communities particularly when faced with the process of economic liberalisation and global and local capital investments into their habitations. For tribal areas also being forest areas with rich mineral deposits and other natural resources, the tribals are frequently targeted to evacuations, denial of compensation and rights to self-government as envisaged in the PESA Act. The state governments have strong vested interests in supporting capital that invests in these areas, than in protecting the interests of tribal communities. The policy process dramatically plays itself out in the case of PESA Act, wherein the actors influencing the policy include the state governments, the various contractor lobbies, the capitalist entrepreneurs, and the different pre-existing line departments such as the Forest Department, which have stakes in protecting their interests. What often results is a non-implementation of decentralisation policy and the marginalisation of the interests of the local tribal communities. The PESA law even when it is implemented is implemented only in a diluted form as various interests described above prevail upon the governments. The policy process vis-à-vis the PESA Act is less researched and makes an interesting case for research from both institutional as well as policy processes point of view. The tribal communities that inhabit the Scheduled Areas happen to be some of the most marginalised communities in the country and as mentioned above the areas that they inhabit consist of rich and myriad kinds of natural resources. These range from timber and medicinal plants to coal reserves and many other varieties of minerals that happen to be raw materials for the major industries [what some social movements in India call as Jal, Jangal aur Zamin (water, forest and land)] issues. Often this process of neglect of tribal interests also results in violent conflicts. While considering decentralisation as a policy process we need to pay attention to these aspects.

Conclusion

But how does the decentralisation policy become ‘iterative’? Firstly, the policy as intended in the beginning, in the form of Acts of Parliament, alters to considerable extent when it reaches the ‘end user’. All the above-mentioned influences go into making the alterations. The policy is only, often, partially realised. But at the same time, on the positive side, the decentralisation process also brings, and has brought already about, considerable change in its wake; particularly in the perception of polity by grassroots people and their role in it. The main features of the decentralisation policy such as affirmative action, timely conduct of elections, delineation of finances through State Finance Commissions, interact with and are influenced by the interests of the political parties, political competition among the parties, the good intentions and activities of NGOs/civil society actors/advocacy groups. All these result in a mixed, if imperfect, realisation of the policy. Feedback from the grassroots level to the Parliament is scant. It is, to begin with, the Parliament which made these Acts possible in the first instance, and in order for the same Parliament to reconsider the policy, and make it more effective, the policy reconsideration may not have happened. However, the main intended beneficiaries of the policy have surely gained at least partially in the process. In that policy process, democracy has to some extent deepened, though ‘economic development and social justice’ still elude the weakest.

In this paper, we have argued that decentralisation policy in India should be seen as a policy process and an interactive and ‘iterative’ process at that. Most of these aspects are well known; what we only attempted here is to present these from the prism of policy processes. How can a public policy process not be ‘interactive’ in a democracy as vibrant as that of India, and that the decentralisation process in India too renders itself to policy processes explanation, perhaps, goes without saying.