The place of data in governance is central and indisputable, yet complex and difficult. There are two parts to the centrality of data in governance: how accessible it is, and how easy is it to ‘read’. For many, many years, the accessibility itself was in deep distress in India. However, since the passing of the Right to Information (RTI) Act in 2005 by the parliament of India, great strides have been made in transparency in governance. While systems still have some way to go before gaining the desired level of openness, there is no question that far more data is available far more easily now than ever before. Does this mean that one can vet decisions taken by governance institutions against ‘data’ that depicts the reality of needs? This is where the record of data ‘use’ becomes difficult to interpret.
The achievements and challenges related to open government data in India have been outlined by the recent report by the Transparency and Accountability Initiative [1]. The report correctly points out that beyond accessibility per se, it is the ‘information’ content of data that needs focus. This is especially true when it comes to decision-making within governance. Post facto accountability is a somewhat meaningless exercise if the decisions whose outcomes are being measured were made in an ad hoc manner.
In some sense, the data does not become ‘open’ until the insights it might potentially contain get revealed, and utilised fruitfully. But, all analysis needs a context, a question to answer, an insight demanded for a particular purpose. So openness is not only a matter of applying methods of data analysis, but also of fixing the context. If both exist, then openness works as a demonstrable principle, well beyond accessibility alone. While data collection might be done in a fairly systematic, timely and unbiased fashion, analysis of the data for meaningful insights and disseminating the results to the concerned decision-makers remains a very large challenge.
So, who sets the context? Who decides what insights to look for? This interesting and very basic question always remains unanswered. Asking questions of a set of data is a highly learned process. It requires experience to ask the right questions for a given dataset and skill to extract the answers. Do elected representatives have the required experience to be able to do this meaningfully? Can they formulate very clearly what kind of structured input they are looking for while making budgetary allocations and policy decisions? Maybe not, but they can articulate their concerns in some way. Is there an agency that can discover the questions implicit in the concerns being expressed by elected representatives and analyze available data accordingly to take the process of debate and consultation forward?
There are very few agencies that might fulfil this function. The first and best placed among them is the bureaucracy. In fact, it is one of their primary functions. They bring both sides to the table: a) they assist in decision-making and are privy to the concerns being expressed by elected representatives and, b) they know the data and have the required skills to analyze it. And yet, this ideal relationship does not yield ideal results, only incremental ones. There can be many reasons for this, but perhaps the most important among them is the fact that in a democratic system, governance is fundamentally political. Elected representatives are not necessarily looking for ‘the best decision’. They are more likely to be looking for ‘the best politically advantageous decision’. This distinction is by no means meant pejoratively. In fact, in an electoral democracy, this is highly desirable. To reflect the concerns of the people they represent is the fundamental duty of the representative. But politicians do not represent monolithic electorates with unitary opinions or demands. In fact, they are frequently juggling contradictory demands from different sections of their constituencies. And so the question of context comes to the fore again. Who sets the context to an input for a decision? And here, traditionally, the bureaucracy bends to the one set by the ruling dispensation, which in turn is governed by a politically calibrated approach to decision-making, hopefully satisfying more than those who are left out. In such a complex and highly charged atmosphere of decision-making, the bureaucracy has limited enthusiasm to present varied analysis and possibilities. This enthusiasm usually diminishes further as one gets closer to the village government, and the ‘distance’ between a largely rural body of elected representatives and a well educated bureaucracy widens so much, that there is not even any pretence of informed decision-making based on accurate data.
The other important agency is the civil society organisation (CSO). The CSO approach is usually to focus on a particular issue, gather evidence in its support and put pressure on the system to decide in its favour. They set the context for themselves, and generally do a thorough job of analyzing data to support it (when they do use it). While this is a straightforward way of using data and its insights, their record of success is not always what one would expect it to be. And again, it is politics that is the un-factored parameter. CSOs frequently work from outside the political system and have limited success in convincing elected decision-makers that this is not only a good decision, but it is also good for them. Listening for the questions implicit in the concerns of the elected representatives and addressing them by giving nuanced input is not a function that CSOs are known to perform.
So openness of data in the specific arena of decision-making in governance remains a challenge. Creating information from data that will shape the debate on any decision is an approach adopted by the higher levels of government rather than the local bodies. However, local governments are far closer to the ground in both terms; assessing what is required, and delivering it. The constituencies of their elected members are smaller and they are more visible to the people than those elected to higher levels of government. Their responsibilities also tend to impact the day-to-day lives of their electorate far more directly, and they can be held accountable for failures in these matters also in a more direct manner. They are also much better placed to give input regarding local conditions to higher levels of government regarding policy decisions, especially for regional development. However, it is precisely at this level of governance that data based inputs become rare. The bureaucracy, local CSOs and the elected members are ignorant of the sheer range of data available to create inputs for local decision-making. Asking questions and creating contexts remains a huge hurdle. The distance between data collection agencies and users at local levels is very large, creating conditions that essentially mean that the data is simply not used, even when this data is open and available.
This paper deals with one concrete example of such a case of lack of data use at the local level. Economic development is an area of governance that is frequently allocated to higher levels of government, while village or city level institutions are viewed as service delivery agencies. However, when it comes to the knowledge of local economies and what ails them, local elected leaders are likely to have a much greater depth of understanding. Can they not be brought into the ambit of consultations and action on regional economic development? If so, how accessible is the existing data? Can it be used to create insightful inputs to a debate at the local level? This paper will present work done to assess the degree to which data can be opened for discussion among local decision-makers. It will also make an attempt to gauge the capacity of institutional mechanisms of democratic political governance to take the process forward.
Planning for social and economic development is an essential function of all governments. India has gone through a tumultuous history of planning. It has seen some radical paradigm shifts in the very ideology and method that goes into deciding the how and what of planning. From being a highly centralised planner till the late 1980’s, India shifted to ‘decentralised’ planning by amending its constitution in 1992 and placing highly local constituencies at the frontlines of planning for many sectors of government action. The weight of expectations from government was thus sought to be shifted to the shoulders of those who were closest to the communities they represented; the elected governments of villages and municipalities. A structure of decentralised ‘panchayats’, or local governance institutions, was created that, ideally, linked the village to the union government. This involved a step-by-step devolution of mechanisms for setting priorities, making decisions, handling finances and governing by participatory means. Planning was made a central tenet of this whole structure and each successive institution in this structure was given specific responsibilities to incorporate plans (or demands) of those in the tier below them in an upward cascading process of convergence that would then organically arrive at a plan for large regions.
The administrative structure of governance in India, viewed from this perspective is seemingly simple.
The governance units in this structure are the GP, Municipality, TP, ZP, State and Union. The trio, ZP, TP and GP are commonly referred to as Panchayati Raj Institutions (PRIs). There is, by and large, a clear separation of functions of each level of governance and this is set down in the constitution. The States and the Union were created and are very well defined since independence. The existence of districts as strong administrative units goes back to the British and has persisted. In fact, the administrative office of the District Collector remains a major power centre and even today the elected ZP is a parallel and much weaker body. Blocks or Taluks were a somewhat late entrant and were minor administrative units. By and large, the real governance and political gainers in this process of decentralisation have been the GPs. In a vastly rural country like India, this is perhaps not surprising. However, the politics of sharing power in the age of decentralisation is very complex, and some of its nuances will be highlighted in this study.
Indian States have distinct governance ‘personalities’ and histories right from pre-colonial times. Since the country was created as a Union and States were given great autonomy of legislation, these differences have persisted. Some States, like Maharashtra and Madhya Pradesh, experimented with decentralised governance, creating elected bodies at the district level and strong GPs. Others stayed highly centralised with the State functioning mainly through the administrative machinery of the District Collector with elected bodies only in the GPs. The political equations between State legislators (Members of Legislative Assembly or MLA) and GP members were shaped by these realities. Since all programme and spending decisions were made by the State, there was a strong top-down nature to governance and State legislators (MLAs) enjoyed great power. However, all this was set to be changed with the passing of the decentralisation amendments.
The 73rd and 74th constitutional amendments [2] in 1992 formalised not only the notion of the GP, municipality, TP and ZP as ‘local self governments’ (LSGs) with elected bodies, they also laid down the functional and financial space for each: to deliver economic development and social justice. There are some key features of the defined ambit of their functions that are worth keeping in mind:
It is the responsibility of each State government to devolve by an act of the State assembly, the required funds, functions and functionaries to enable all LSGs to attain legal status and perform. This has led to an uneven record of decentralisation in India. Some States have done better than others, but none have reached the level of LSG empowerment and functioning envisioned in the constitution.
Most LSG functions are service oriented, such as drinking water, sanitation, watershed development, education, health, and so on.
LSGs have limited financial resources. GPs and municipalities have some powers to tax and levy service charges. However, both TPs and ZPs have scant sources of their own revenue. For most part, all LSGs depend heavily on State governments for devolution of funds to meet the expenses of their programmes.
The functional relationship between States and LSGs is mediated first and foremost by the LSG bureaucracy, whose appointment is entirely in the State ambit. The second arm of the State present in LSGs constitutes all departments (or ministries) of the State that function at LSG level. So, education, health, irrigation, public works, and so on, all have functions at the LSG level that are independent of the LSG and controlled by the relevant State minister.
Elected representatives for GP, TP, ZP and State government share voting constituencies, and State assembly members have ex officio positions on important bodies such as the District Planning Committee. The functions of the four levels of governance institutions are different, but they are meant to dovetail. Only the State has legislative powers, and the power to devolve State revenue for LSGs.
The combined administrative, political and financial footprint of the State is very large in LSG jurisdictions.
These developments created two shifts: one in the functional and financial arena and the other in the political. The power equations between the MLAs and GP (and now TP and ZP) members shifted, with a perceived dilution of the clout of the former. With deeply overlapping voting constituencies, much had to be shared and negotiated, a new experience that the MLAs resented [3]. The unresolved struggle to come to terms with this upending of the old political power hierarchy is seen in this study also. With the creation of PRIs left to State assemblies, this shift in the power balance is one of the main reasons why it took States a long time to operationalise PRIs and remains an on-going process.
Since planning was a highly centralised function at the State and Union levels, it has taken even longer to bring it into the ambit of the decentralised governance regime. It was only in 2006 that the Planning Commission of India created the first direction for States to create district plans and has since come up with the Manual for Integrated District Planning [4] giving details of the planning process to be followed by all LSGs. It is comprehensive on process, data management and integration. It emphasizes visioning, participatory methods and social audits. Many civil society organizations are taking part in awareness building, training and demonstration regarding the implementation of this whole exercise. Vast numbers of administrators, elected representatives and citizens have come into their ambit and have benefited tremendously [5].
Planning in this system of governance, ideally, follows an upward aggregation path. Each GP and Municipality makes a plan for its own needs, based on public consultations and available resources. All GP plans are aggregated at the TP level. The ZP takes into account all TP plans and adds items that are usually pan-ZP and cannot be split, such as long distance roads or irrigation projects. The State government then takes into account ZP plans, adds items that are pan-ZP, along with those prepared by each of its ministries for functions that only the State could perform. It then balances its financial resources among all these needs and creates the State plan.
However, the reality is quite different. Planning has historically been an administrative bastion, with only a few State level ministers overseeing the process. It remains so today, even in the decentralised regime. This is formally so. The ZP has a planning department that is given the responsibility to plan according to the various directions of the Planning Commission. This plan is supposed to be based on the wish lists presented by the ZP members from each of their constituencies and the GP and TP plans. Even if one grants that the district planning office is efficient and reflects these wish lists to the best possible extent and the resultant plan is passed by the elected body, there is a catch. The ZP plans have to be ‘approved’ by the State. This is not an arbitrary process in principle. The State must prepare a comprehensive plan that includes PRI plans, and its own cross-State programmes. In effect, there is a re-casting of the entire plethora of plans. And not everybody can get everything they want. The matter is down to a balance between what must be done, and what can be negotiated. This is an essentially political arena, where elected leaders from different regions compete for a limited resource i.e., State funds.
In this crucial process, the ZP is represented at the State level by the ZP chief administrator and not by the elected president. This is true of all States. From this point onward, the elected body has no control over what happens to the plan that they have passed [6]. It is at this crucial juncture that the disconnect between the ZP members from the MLAs proves fatal. With no political weight to be exerted at the State level on behalf of the district working to a single plan, they are left with having to approach bureaucrats in various ministries to grant them their wishes.
There is a tendency among civil society organisations who work with the PRIs to view the process of decentralised governance as an essentially apolitical one: if ‘good’ technology, intentions, methods are deployed, the principle of subsidiarity will work at the village and district levels and people’s concerns will be reflected in the plans and programmes. They consider the State bureaucracy to be more of an ally in the battle for the rights of the PRIs, than MLAs. However, the very act of the creation of PRIs under the 73rd and 74th CAAs is essentially a demonstration of political will at the State level. Governance at the State level is deeply political, and the upward integration of PRI plans cannot be set outside this ambit. While the bureaucracy can mediate, it does so at will, and can only accommodate such approaches within the limits set by State level political agenda for governance.
So, whither decentralisation? It’s still a patchy business in India, with many battles being fought on many fronts, most of them fundamentally for greater political space at all levels of institutional governance. It is this that makes the institutional evolution of this idea so fascinating to watch. As local players learn more with each passing year about how the system works - and does not work - for them, there is upward pressure created from below to correct the perceived imbalance, that generates some give and some push back. And so the cycle continues. The change is slow and incremental, but then in a vast and messy democracy like India, perhaps it is best to let loose a large idea and allow it time to become something unique within the existing institutional structures.
This work began as a way of testing the ability of data generated by various government agencies to give meaningful, detailed input to the process of district (ZP) planning. The study did not address planning at any other local level, such as GP or TP, but was restricted to district planning at the ZP level. There are many processes of district planning being put in place over the past few years, with their methods, detail and focus becoming more diverse since the passing of the 73rd and 74th CAAs. The focus of this study was economic planning.
There are two parts to assessing the database for economic planning: 1) existence of open, accessible, usable, relevant data and 2) the engagement of the stakeholders, such as elected ZP and State assembly members, and administrative ZP staff, in the process of economic planning.
There must be some factual information about the district and its economy on which planning for further development can be based. Identifying past and present trends in the performance of different sectors of the economy is necessary to make an assessment of what will work in the future. Many lessons can be learnt from such analysis. Regional imbalances can be identified in far greater detail. This data can be used by many levels of governance to further their agendas for development. The data used in this study, for instance, was a census of enterprises. However, its analysis was greatly enriched by discussing it with the local stakeholders. Why did various factors of the local economy change the way they did? This is not a question that the data can answer, but the officials could. In fact, unless such data is discussed thoroughly with local populations, it is difficult to see how the underlying processes of change can ever be fully understood, much less used as the basis of planning for the future.
The aim of this study was to verify whether there is already sufficient data available in sufficient detail within various government agencies to make for a usable and accessible database that can be of interest to all levels of government, from the State Government to the GP. The second aim was to create a way of presenting this data to various stakeholders to assess its usefulness on the one hand, and their ability to absorb the information on the other.
The first phase of this study used data from the Population Census of 1991 and 2001 done by Census of India and the Economic Census of 1998 and 2005 done by the Central Statistical Organisation (CSO) [7]. Data on state and district-wise agricultural productivity was taken from the Directorate of Economics and Statistics, Department of Agriculture and Cooperation, Ministry of Agriculture, Government of India.
The location of the study is Madhubani district of Bihar. The study was conducted at the ZP level.
When it comes to using data to the smallest geographical unit like the village, there are not too many choices. The population census conducted once every ten years is one major source. State governments have a system of data collection for health, education and a few other social services for which data is maintained on an annual basis at the village and town level. It is contested [8], many times, but it exists. By and large, in many States, the formats are consistent and the data is usable. Finances of LSGs are available, but their usefulness is quite limited, since the reporting system followed is very far from intuitive and at times positively opaque. Much of this data can be obtained, ideally, with an application under the Right to Information (RTI) Act.
Some States do post such data on their official websites, but not for every village. Very few GPs have websites. Most ZPs have websites which contain some interesting data. For example, district websites give the entire database of village census to identify households ‘below the poverty line’ (BPL) which are eligible for subsidised food, housing and other services. These are available by household and are updated whenever a new census takes place, generally once every five years or so. Many small and large municipalities have websites, and in many States they post some data online. While the manual on district planning advocates the use of GIS and map technologies, this has not been implemented yet.
The National Informatics Centre (NIC) [9] is the Union agency created for e-governance solutions, with offices in all district headquarters, tasked with creating, maintaining, and upgrading all government websites. While their task is far from complete, they can be credited with going a long way towards that goal. They have created a GIS based Planning Atlas for a few states[10]. However, these are quite limited in their imagination of the planning process, and it is not clear how much they are actually used even at the State level.
It must be understood, of course, that data is ultimately political. It is also used differently by the political class (from GP to State), and by the bureaucracy. The basic power imbalance and adversarial relationship between GP/ZP members and the MLAs has been outlined above. The bureaucracy, on the other hand, is highly integrated from the highest State level right down to the GP secretary [11], and does close ranks. Data is a weapon that is much sharper in the hands of the bureaucracy than in those of the politicians and has always created a fundamental imbalance between the elected and administrative wings of governance at all levels. But by and large there is less collusion within the political hierarchy than there is within the bureaucratic one in the matter of biased or selective use of data, or even outright data manipulation. With low education levels, inability to understand the (very important) detail of how data is collected, analysed and used for planning, and limited tenures, politicians at all levels of governance are at a basic disadvantage in asserting themselves in a tussle with the bureaucracy[12].
Technology and data have increasingly become synonymous in the governance arena. The use of computers has penetrated into governance in the developed states much more than in backward ones, but the momentum has been built and resources made available to complete this process down to the GP level. Availability of electricity, skilled users and technical support remain the main stumbling blocks in this process, especially at the GP level. However, all ZPs have computers for all kinds of record keeping, and also internet access. Even so, it is clear to all stakeholders that the introduction of new technology alone is not going to make processes of governance more accessible, especially in matters of planning. Its use must become much more inclusive and intuitive for primary decision-makers, the elected representatives. As of now, it is yet another layer of separation between them and the bureaucracy. This work attempted to demonstrate that, in fact, technology can be used intuitively to provoke constructive thinking in the planning process.
In fact, the term ‘open data’ can be applied to the process by which data is made usable through its analysis and interpretation. It is made open to learning and debate. Raw, unit level census data can hardly be considered ‘open’ from the point of view of any elected representatives, even though it is easily accessible. In principle, it is the function of the planning machinery, and civil society organisations, to be the transformative intermediaries that ‘open’ such data for them.
This paper deals with data that is not of great interest to anybody at the local or State level, is the result of periodic census exercises, and for these reasons, highly unlikely to be subjected to manipulation. It also reflects the ‘economy’, for which no local elected body takes clear ‘responsibility’, unlike schools or health centres. Given this political tendency, the administration was quite sanguine that there would be no difficulties created by the elected body based on this unwanted information offensive.
Most data on the economy is collected in the form of sample surveys by Union agencies, which are rarely, if ever, statistically significant below district level. As a result, sub-district, regional spreads of various economic indicators are very difficult to obtain. This is also because State governments have never taken the trouble to institute State wide data collection mechanisms that match the methods used by Union agencies, but go below district level for their own planning purposes. This is ironic, because the Union agencies use State and district level offices to collect their data. So there has always been a large data gap for State level use, and the States alone are responsible for it. This also reflects the highly centralised economic planning process that has been the tradition in India.
The Economic Census of enterprises conducted by the CSO is really the only means available to do a detailed regional, sub-district study of the non-farm economy in India. It can be broadly supported by occasional surveys of specific sectors done by the National Sample Survey Organisation (NSSO), and the Annual Surveys of Industries done by the CSO for large industry.
Both population census and economic census unit level data is available for purchase in machine readable format for more recent years. There are no restrictions on its use. State maps are published by the Census of India after every census and are available to the level of the village boundary. These are also available for purchase. All transactions are reasonably straightforward, and the CSO offers continued long distance support for clarifications and missing bits long after purchase.
The real question to ask in the context of this study is: was this data being used within the planning process at the district level in Madhubani? And if not, could its possible use be demonstrated? After all, data is not just about numbers filled out in a form, but what they tell one about the quality of life of people in the country. Data needs to be interpreted from as local a perspective as possible to get the most out it.
Since village census is done for BPL or other beneficiary identification purposes once in a while, the dependence on population census is not very heavy. The data that is not collected locally periodically, but is available in the decadal census, such as worker data, is not used. In fact, there is very little data used on the economy. For example, a powerful indicator of the state of employment in the district, agricultural and otherwise, is the number of marginal workers, who do not work for more than 6 months of the year. Even the simple changes in the number of marginal or full time (main) agricultural and non-agricultural workers over a decade are not used to understand how employment opportunities are shifting and why this may be happening.
Madhubani ZP has a Vision 2020 [13] document on its website that gives an overview of its status, a SWOT analysis and suggestions for development over the next 20 years. This was prepared as a part of the mandated planning process by a reputed civil society organisation on its behalf, a common practice. However, it is more an enumeration of some of the available data, than its in-depth and localised analysis, and consequent suggestions for possible paths to be taken for development. This document also did not use the Economic Census data.
The service sectors of governments, especially subsidised ones, such as health, education, housing, public distribution of food, water, electricity, roads, receive the most attention in participatory planning processes. This is logical and easy to understand. These are day-to-day issues that impact the immediate quality of life very significantly. For most of rural India dependent on agriculture, irrigation, availability of seeds and fertilisers, and a support structure for failed crops are also imperative factors. The track record of the provision of these services is so poor, in such vast swathes of India, that a fierce focus on them from local stakeholders is inevitable and highly desirable. And these concerns are repeatedly reflected in the plans that come from the GPs.
However, who is to look at long term issues of generating employment and planning for an economy that must take into account the stagnation in agriculture and the lack of employment opportunity? Which agency at the district level is responsible for this analysis and its propagation among key decision-makers within the planning process? In the case of Madhubani, not one agency was willing to lay claim to this task. Employment generation and the rethinking of local economies were contended to be the task of ‘State governments’, not the ZP or the GP.
As a result, data on these issues, readily available from sister agencies, was never used. In some sense, data use at local levels is more reflective of issues that are traditionally considered immediate, quality of life concerns. While employment, wage and the long term growth of the local economy are also factors that will greatly impact quality of life, they have never been cast as such in the various manuals and training courses on planning, and so receive little attention. Even though the constitutional amendments put the responsibility of providing ‘economic development and social justice’ onto LSGs, this is not the case in practice.
Given this state of planning, it was worth putting together the most broad-based data on the economy, to see how it could be used to create intuitive information and disseminate it. Was such ‘opening’ possible, and how did stakeholders react to it?
The following sections give some of the more successful lines of analysis that were used in this study, and their reception at various levels of local governance.
Let us start with an overview of Madhubani district. One of the northern districts of the northern state of Bihar, with a border with Nepal, this is one of the most backward districts in the country. It is spread over 3501 sq kms, and has 21 blocks, 399 GPs with 1034 villages. It has more than 10 large and small rivers and numerous streams running north-south, bringing flood waters from the Himalayan foothills in Nepal. The district is dotted with over 800 large and small ponds.
The population of the district was 3,575,381 in 2001. Male literacy was about 57%, and female 26%. While 323 villages had domestic electricity connections, only 232 had some medical facility. Paved approach roads connected to 547 villages, and only 164 villages had bus services. Infant mortality was 62 per thousand births.
A very backward district on all fronts, Madhubani is a planner’s challenge. Poor infrastructure, flooding, and a very long history of poverty give it a historical disadvantage. However, Bihar as a whole has a strong political tradition, with some remarkable leaders having emerged from these regions. While politics remains dominated by caste divisions, Madhubani exhibited a strong understanding of the spirit behind decentralisation, in no small measure due to the efforts of civil society organisations. The ‘rights’ of LSGs and ‘duties’ of the State were regularly invoked by ZP members to convey the lack of real empowerment of LSGs. The relationship between elected ZP members and the bureaucracy remains a balancing act, with each side claiming superiority over the other to get their way. The president of the 56 member ZP was a young woman from a backward caste, who had won her seat in a non-reserved category.
Meaningful inputs for economic planning for such a district had to take into consideration not just the data but also methods of dissemination that would appeal to a very diverse audience.
Data sorting is a fairly technical exercise. To bring data from two different agencies, Census of India and CSO, in line with each other poses a number of difficulties. Starting from spellings of village names to the location codes used, sorting them out requires painstaking effort, repeated runs and constant cross checking for accuracy. However, while highly time consuming, for those who understand the data structure, this is the smaller hurdle to cross.
The challenge is to come up with lines of analysis that accurately identify the major events and trends in the history and geography of the economy of the district. They must logically link the demography and economy down to the village level, and present scenarios for what can be done in the future. Intimate familiarity with the obvious patterns in the data is important. Field visits, while very necessary, make the most sense when the overall data is already in place. This study certainly benefited tremendously in this respect by travel within the district, conversations with civil society organisations and their local workers and elected representatives such as members of State legislative assembly (MLA), ZP and GP members. They all had some input in pointing at different areas of concern, making for a much more vetted analysis.
Apart from the data itself, an extremely useful tool was introduced to the study, that of digital maps of the district. Census of India prepares maps of the district every 10 years, down to the village boundaries. These maps also have a great deal of other information: locations of schools, health centres, post offices, banks, rivers and streams, roads, railway lines, and so on. Digitising these maps [14] and linking them to the data about the economy proved to be the final link in creating a remarkably clear picture of the history and geography of the development of the district from 1991 to 2005.
All maps given in this document have been created using data for villages only, not for blocks. Block boundaries are marked for ease of visual identification. There are no averages in the maps.
There are three lines of analysis given here about the economy of Madhubani district. This is not an exhaustive account of all the work done, but more illustrative of what emerged as important issues for this district.
1) Agriculture
2) Urbanisation
3) Non-farm employment
These areas of interest have been chosen not only on the basis of their importance for an analysis of the economy of Madhubani district, but also from the point of view of similar concerns operating in large parts of India.
At the time of the 2001 population census, Madhubani was a highly rural, agricultural district, with 96 percent of its population living in villages and female literacy at 26 percent. Rice is the main crop and staple, with rabbi wheat a close second.
Many large and small rivers run through this district. They flood every year and inundate large areas. Embankments have been built over some stretches to contain floods, but they seem to have only a partial mitigating effect. While ground water depths are fairly shallow, there is still a demand for irrigation for the rabbi crops. The Kosi canal cuts through the district east-west and carries water during the monsoon, but very little later.
Agriculture is the main employer in the district. Table 1 gives some basic numbers.
| Basic demography | 1991 | 2001 | Decadal growth rate |
| Total population | 2,832,024 | 3,575,281 | 26.24 |
| Main Cultivators | 372,262 | 291,565 | -21.68 |
| Main Agricultural labour | 370,731 | 416,627 | 12.38 |
| Total main in agriculture | 742,993 | 708,192 | -4.68 |
| Main others (non-farm) | 96,185 | 169,220 | 75.93 |
| Total Main Workers | 839,178 | 877,412 | 4.56 |
| Marginal male | 4,435 | 133,381 | 2907.46 |
| Marginal female | 76,639 | 216,930 | 183.05 |
| Marginal in agriculture | NA | 311,807 | NA |
| Total Marginal workers | 81,074 | 350,311 | 332.09 |
| Total workers | 920,252 | 1,227,723 | 33.41 |
Marginal
workers: those who have worked for less than 6 months of the
year
Source: Population Census of 1991, 2001
A very high rate of population growth has not been accompanied by rising employment in agriculture. While nearly 81 percent of main workers are engaged in agriculture in 2001, this number has seen a small decline since 1991, when about 88 percent were main agricultural workers. Cultivators have reduced substantially. However, marginal workers have seen a sharp rise, especially male. Basically, agriculture has seen great stagnation over these 10 years and the rise in total workers is largely driven by the rise in marginal workers, 90 percent of whom were in agriculture in 2001. Non-farm employment has seen a rise, but the numbers have remained small. A crisis of employment has gradually set in over the decade of the 1990’s.
These numbers in themselves are sufficient to give an overall idea of what the economic scenario of this district has been. However, using maps adds a further degree of refinement, giving a clear picture of the intra-district imbalances.
It becomes clear from Map 1 that the central and eastern parts of Madhubani district have suffered the most in terms of the rise of marginalisation of labour. Laukahi, Khajauli, Kaluahi, Ladania, Andhratharhi, Lakhnaur, Bisfi and Babu Barhi blocks have seen a large proportion of their villages having high rates of marginal workers. Western blocks such as Benipatti and Madhwapur show marginalisation to a smaller degree.
So not only is it possible to understand the overall employment conditions in the district, it becomes possible to understand which geographical areas have suffered the most set backs and where poverty levels are likely to be high. Targeting such areas preferentially with programmes like the National Rural Employment Guarantee Act then becomes feasible and defensible.
There is no reason why GPs should not use such data to demand a commensurate level of assistance. It is also possible to devise schemes for boosting agricultural productivity in these areas, making sure agricultural extension work is made more targeted.
Embankments or not, if floods are inevitable, causing a large amount of down time in agricultural activities, then methods of cropping have to be devised to make sure that the rabbi and summer crops yield enough to make up for it, both in terms of days of employment and productivity. Table 2 gives the position of Madhubani district in rice productivity in comparison with other states.
It can be seen that for the most part Madhubani district produces less rice per hectare than the Bihar average, and lags far behind many other rice producing states. The wheat statistics shows similar trends.
| Year | AP | Bihar | Punjab | UP | WB | Madhubani |
| 1996-97 | 2.6 | 1.44 | 3.4 | 2.1 | 2.2 | NA |
| 1997-98 | 2.4 | 1.4 | 3.47 | 2.2 | 2.2 | NA |
| 1998-99 | 2.8 | 1.33 | 3.15 | 1.9 | 2.3 | NA |
| 1999-2000 | 2.7 | 1.45 | 3.35 | 2.2 | 2.2 | 1.36 |
| 2000-01 | 2.9 | 1.49 | 3.51 | 2 | 2.3 | 1.14 |
| 2001-02 | 3 | 1.46 | 3.54 | 2.1 | 2.5 | 0.96 |
| 2002-03 | 2.6 | 1.42 | 3.51 | 1.8 | 2.5 | 0.77 |
| 2003-04 | 3 | 1.52 | 3.69 | 2.2 | 2.5 | 1 |
| 2004-05 | 3.1 | 0.79 | 3.94 | 1.8 | 2.6 | 0.28 |
| 2005-06 | 2.9 | 1.07 | 3.86 | 2 | 2.5 | 0.89 |
Source: Directorate of Economics and Statistics, Department of Agriculture and Cooperation, Ministry of Agriculture, Government of India
With such low, and declining, productivity it is not surprising that there is large scale marginalisation of labour in agriculture. The sector has been able to absorb only marginal workers after 1991. Population growth is too large to be sustained by such trends. The distress in agriculture is very deep, indeed, and the following case studies will show its impact on other sectors of the economy.
Urbanisation as a phenomenon arouses much interest among economic planners in particular. It is assumed that economic progress can be measured in terms of the degree of urbanisation, since most of the GDP is generated by non-agricultural enterprises, which are largely urban based or give rise to urbanisation in their vicinity.
The data of population census of 1991 and 2001 was used to map the patterns of urbanisation in Madhubani, with the following definitions:
Urban: Workers in agriculture < 50 percent of total workers
Semi-urban: Workers in agriculture between 50 and 65 percent of total workers
Semi-rural: Workers in agriculture between 65 and 80 percent of total workers
Rural: Workers in agriculture more than 80 percent of total workers
Map 2 shows that urbanisation, as defined above, has taken place preferentially in west-central areas of the district, largely around the town of Madhubani, the district headquarters.
It can be seen very clearly from Map 3 that most urbanisation (red, pink and green), or an increase in non-farm employment, has taken place in areas where there is the most dense road network, along with a railway line. Most of the eastern and some of the northern blocks exhibit very little urbanisation and have extremely few roads.
It will be useful to remember that the most urbanised area of the district encompasses Madhubani, Pandaul, Rajnagar and Kaluahi blocks, with a lesser amount in Jhanjharpur and Lakhnaur. Bisfi block shows both, large scale marginalisation of workers in agriculture, and has become semi-rural with an increase in non-farm employment. This has been mainly due to the presence of a railway line and proximity to a road link from Darbhanga to the south.
It is also easy to see that while a great deal of marginalisation of labour has taken place in the eastern blocks, they still do not see a viable option in non-farm enterprise coming into play. The economy is so poor that non-farm employment has not taken hold to any degree at all. The absence of roads is an added handicap.
This gives even further impetus to the planning needs of programmes like NREGA. These areas are in multiple distress conditions, and should be targeted on high priority basis for employment generation and creation of public assets. Since this data is available at the village level, identification would be very easy.
The Economic Census done by the Central Statistical Organisation for the years 1998 and 2005 is invaluable in tracking patterns in the non-farm enterprise sector. The Economic Census is a census of enterprises, giving various details about ownership and employment. Again, data is available at the village level. The basic data for these two years for the district as a whole is summarised in Table 3.
| Sector | Workers 1998 | Workers 2005 | Enterprises 1998 | Enterprises 2005 | % Rise Workers | % Rise Enterprises |
| Primary sector | 4414 | 2144 | 2982 | 1173 | -51.43 | -60.66 |
| Manufacture and repair | 15783 | 17770 | 6987 | 9739 | 12.59 | 39.39 |
| Wholesale and retail trade | 20575 | 34628 | 13755 | 21799 | 68.3 | 58.48 |
| Hotels and restaurants | 3228 | 5275 | 1719 | 2988 | 63.41 | 73.82 |
| Public admin, education, health | 22247 | 14076 | 7478 | 5570 | -36.73 | -25.51 |
| Others | 3079 | 3133 | 996 | 1402 | 1.75 | 40.76 |
| Total | 69326 | 77026 | 33917 | 42671 | 11.11 | 25.81 |
The trends in non-farm enterprise employment are as alarming as those in agriculture. The primacy of employment held by public administration, health and education services in 1998 has lost ground very heavily in just 7 years. Instead, by 2005 the dominant sector is wholesale and retail trade (more than 90 percent of this is just retail trade). Manufacture and repair see a small increase but in absolute numbers they remain small. Even the primary sector, which includes animal husbandry and fishing, has lost ground in spite of the fact, that there are more than 800 large and small ponds, and many rivers in the district. Fresh water fish from Andhra Pradesh is sold in local markets at the same price as local fish. The number of buffaloes has actually declined. The one makhana (seed of pond vines, used in sweet making) plant in Jhanjharpur was lying vacant, and production was still done on an individual scale. Fishing and makhana production were mired in highly inefficient and near defunct cooperatives.
A more detailed look at each of these sectors reveals a pattern of rapid withdrawal of the organised sector as a whole. Large rice mills, spinning mills, handloom and power loom units have shut down. The most job losses have been in public administration, at nearly 43 percent, followed by education at about 35 percent.
In 1998 about 76 percent of all enterprises were own account with 53 percent of all workers, by 2005 this proportion was 91 percent own account enterprises with 86 percent of all workers. The hiring capacity of the non-farm enterprise sector drastically reduced over this period. In 1998 the proportion of hired workers to total workers was 41 percent, but by 2005 this proportion was down to just 29 percent.
Did different areas of the district display different kinds of behaviour in the non-farm sector as well? Was there a geographical pattern to the decline seen in the Economic Census data? Was it indeed the case that as agriculture went into distress, it was the urban areas that held up the non-farm sector to whatever degree possible?
To check this, each village and town in the Economic Census was marked by the urbanisation category of 2001 as outlined in Case 2 above. This effectively linked the two data sets, creating a composite picture of the economy of the district. Table 4 gives the spread of hired workers for 1998 and 2005 by urbanisation category.
It becomes clear from Table 4 that the biggest loss in hired employment was suffered by urban areas. This is partially due to the fact that most offices of public administration are located in urban centres. But not all losses in urban areas are in this sector. Decline in small and medium manufacturing has also played a large role. The only area to see a small growth in hired workers was, in fact, rural!
| Region | Hired workers 1998 | Hired workers 2005 | % Rise |
| Urban | 8268 | 3962 | -52.08 |
| Semi-urban | 5102 | 3588 | -29.67 |
| Semi-rural | 5406 | 4125 | -23.7 |
| Rural | 9499 | 10535 | 10.91 |
| Total | 28275 | 22210 | -21.45 |
Map 4 of hired workers showed this phenomenon in greater detail, giving a visual image of the withdrawal of the non-farm sector and its employment capacity.
The most marked withdrawal of high hiring capacity (more than 60 percent of all enterprise workers in a village are hired) in non-farm enterprises has been from the most urbanised areas of Madhubani, Pandaul, Rajnagar and Kaluahi blocks. Similar losses can be seen in Jhanjharpur, Benipatti, Ghoghardiha and Laukaha.
The proliferation of own account retail trade has been a reaction to tremendous distress in agriculture and a failed organised sector. It cannot and should not be counted in the ‘contribution to GDP’ logic attributed to the non-farm sector. For want of any other sustained avenue of work, people are putting up shops in the hope that the daily sale will take care of daily food needs. However, it is highly unlikely that this is the case. Enquiries in the district indicate that the daily sale proceeds of such establishments amount to about 35 rupees. Urbanisation is not always an indication of ‘economic progress’ and panacea for the ills of agriculture.
Complex, yet highly localised analysis is possible with data available from various government agencies, without resorting to primary data collection and other time consuming and debatable activities. If the goal of the planning exercise is clearly defined, lines of analysis can be thought of that give the most relevant input. Maps add tremendous fresh insight to the numbers, making for a powerful planning and monitoring tool. Targeting for delivery of social sector services, especially in States like Bihar, with large rural hinterlands and poor connectivity, can improve manifold with a tool like this.
This kind of analysis is best presented in small groups, rather than to large audiences of largely uninterested members of governance institutions. Tabular data is hard to explain, but the maps made lasting impressions, and retention of key trends and conclusions was much better with maps.
It is also necessary to be ready with some idea of how the difficulties of the situation presented can be remedied. It need not be a ‘plan’ in the formal sense of the word, but it must point the way to action. This could include institutional functioning, utilisation of existing programmes and funds, collaboration with others, and even entirely new ideas. It is only then that the discussion becomes animated and participatory.
A large proportion of this analysis was presented to the members of the District Planning Committee (DPC) of Madhubani zilla panchayat and to the members of State assembly (MLAs) from the district. Some MLAs are official member of the DPC.
There were very different responses from the two sets. The DPC members basically thought that giving them an analysis is not sufficient, but a plan must also be provided for what should be done. Further, the agency that makes the plan must also ensure that the State government then funds it. They knew the state of their economy in a general way, but were clearly unwilling, and perhaps unable, to engage in a discussion about their role in improving it.
This is not hard to understand. The ZP as an institution has really never planned for anything ever since it came into existence. More recent efforts to get them to plan have resulted in various NGOs doing the planning on their behalf, in the absence of even rudimentary ideas about the whole concept. What has happened, though, is that they have suddenly become highly aware of ‘their rights’ under the 73rd constitutional amendment. They are also now aware that operationally, they do not seem to have many of these rights. Most interaction with DPC members frequently revolves around the demand: get us our rights. At the same time there is no demonstrable ability regarding any structured thinking about planning. All they can do is to prepare somewhat disconnected and long winded wish lists, demanding that the State and Union governments fund them.
While they are aware that the State does not wish to give them their rights, they still expect the State to fund their wish lists. To expect both funds and rights from the same entity in a political setting gives rise to confusion about ones own ‘duty’ to act. Sometimes they do, and sometimes they do not. For most part, they do nothing, justifying their lack of action by their lack of ‘rights’.
Very few DPC members had a clear idea of how lack of rights translates into the lack of funds for the ZP. Also to resolve this conundrum they need to work in concert with their MLAs, who have the mandate in both arenas, legislative devolution of rights and allocation of funds. But in the entire time that this DPC had been power, there had not been a single meeting of the DPC with the MLAs. They have nothing to do with each other. The tremendous confusion about their respective roles was clear.
For one meeting with the MLAs to present this data, the ZP president and a few other DPC members were invited. The MLA from Madhubani said this was the first time he was talking to the president about the district as a whole. The president said later that this was the first time that she had been in the same room with most of the MLAs, and had never discussed any plan for the district with any of them. Some DPC members said they had never met any MLA.
In some sense, the whole process of doing the analysis and attempting to present it to the ZP and the MLAs has led to a much better understanding of their respective roles and vast gaps in institutional functioning and connectedness.
The ZP president did understand the motivation behind this study. She showcased the entrepreneurship in the district by pointing at the case of carpet weavers in Basopatti block. These people had been trained in government run initiatives for skill development in carpet weaving. They had then got jobs in the Bhadohi area of Uttar Pradesh. After working there for some years, they wanted to come back to their villages and set up their own units. However, the major hurdle was lack of capital to set up the weaving machines and other ancillary units. They could only work at daily wage level for their employers in UP, but could not do the crucial value addition locally that would have pushed up incomes and created much more employment. Bihar was still not considered a safe destination for even this kind of investment.
The MLAs, however, did not outright reject the idea that planning for the economy fell within their mandate. In fact, they seemed to have a much clearer idea of what they could and could not do. They also reacted very positively to suggestions about how their ability to deliver could be improved. They did have a much longer line of sight in governance, far better than any ZP member, understood the issues quickly and clearly, and were willing to engage in extended dialogue about remedies. It was clear that they are the only realistic allies that the ZP has in its struggle with the State about funds and rights. But there is absolutely no intercourse between them.
The MLAs from Pandaul, Babu Barhi and Kaluahi were quite perturbed by the withdrawal of small and medium organised industry from their areas as shown by the analysis. Pandaul’s small but well developed industrial estate was a ghost town. One major problem often repeated across the district was the lack of electric power. However, some MLA’s were of the opinion that this was something that could be dealt with locally. Yes, there were severe shortages, but something could be made available to industry if it was seen to be functioning. At that point in time, there was no such incentive at work. In fact, most of them insisted that the reasons for the failure of industry was not the lack of power, but other issues, such as competition from Bangladesh, poor labour management, and in some cases, straightforward corruption. The last was held responsible for the demise of all 3 sugar mills in the district, and with them, the demise of highly profitable sugarcane cultivation.
During discussions with MLA’s, the opinion was expressed that it was not sufficient to show the details of Madhubani, but to also show how other places in India were performing, so that they get an idea of what a better developed area looked like in terms of diversity and density of enterprises along with their hiring capacity. They felt isolated in terms of being able to imagine lines along which they could think about pushing sectors of their own economy. This was a reasonable demand. In the next round of interactions, they were shown a comparison between the states of Bihar and Andhra Pradesh.
This analysis is done from the point of view of giving ideas to State and local governments in Bihar about the possible routes they can take to target areas of the economy that can most likely have far reaching impacts on growth. It attempts to take some lessons from another state, Andhra Pradesh, to compare and contrast what has worked and why.
Sources such as the state economic surveys give GDP figures and an overall sense of which sectors of the economy have contributed to what degree. However, the focus here is on the issue of employment, rather than GDP. In India, it is employment, the availability of work for a reasonable period of time and at a reasonable wage, that is the most sustainable path to poverty reduction. However, finding ways of achieving high employment rates is not simple. It is for this reason that Andhra Pradesh was chosen as a comparison, from among all the States that the MLAs have visited. A diverse economy, but with a strong agricultural sector, it presents lessons for how the spread of employment has taken place, with the interplay between the agricultural and non-agricultural sectors playing a pivotal role.
Let us start by taking a look at some relevant demographic figures for the two states from the 2001 population census as shown in Table 5.
| 2001 | Population | Workers | Persons per worker | % Marginal Workers | % Agricultural Workers | |
| Andhra Pradesh | Total | 76,210,007 | 34,893,859 | 2.18 | 16.77 | 62.16 |
| Rural | 55,401,067 | 28,172,888 | 1.97 | 18.44 | 75.04 | |
| Urban | 20,808,940 | 6,720,971 | 3.1 | 9.79 | 8.18 | |
| Bihar | Total | 82,998,509 | 27,974,606 | 2.97 | 24.74 | 77.25 |
| Rural | 74,316,709 | 25,752,569 | 2.89 | 25.78 | 82.4 | |
| Urban | 8,681,800 | 2,222,037 | 3.91 | 12.69 | 17.58 |
Out of a total population of 76 million, nearly 35 million people in AP were workers. In Bihar, out of a nearly 83 million population, only about 28 million were working. The number of persons per worker statistic makes this clear. In rural AP, there are 1.97 persons dependent on one worker. In Bihar this number is 2.89. This shows the degree to which Bihar suffers from unemployment as compared to AP. Even if the basic wage rate was to be the same in the two states, in AP about 2 people will depend on that wage, while in Bihar it will need to support 3 people.
The crisis in employment in Bihar becomes clearer by looking at the figures for the proportion of marginal workers (those who have worked less than 6 months in the year). With a much larger workforce, AP has only about 17 percent marginal workers, while in Bihar their proportion is nearly 25 percent. So not only are a vast number of people unemployed in Bihar, among those who are working, there are large numbers who are not working even half the year. Bihar is projected to have a population of about 96 million by 2010, and if these basic proportions have continued through the past decade, the size of the unemployment problem in Bihar will be enormous.
That Bihar was still heavily dependent on agriculture as its main employer in 2001 is clear from the percentage of agricultural workers in the total. Agriculture supported about 82 percent of workers in rural Bihar, while in rural AP 75 percent workers were engaged in it. It can be said that the economy has diversified much more in AP, creating employment in areas other than agriculture.
It is instructive to see at this point what the agricultural productivity trends have been over the past several years in the two states to understand the capacity of this mainstay of employment.
| Rice productivity | Season | kg/hectare |
| Andhra Pradesh | Kharif | 2555.7 |
| Rabi | 3331.6 | |
| Total | 2800.4 | |
| Bihar | Autumn | 1139.4 |
| Winter | 1359.7 | |
| Kharif | 1325.2 | |
| Summer | 1738.9 | |
| Total | 1337.4 |
Table 6 shows that Bihar lags far behind AP in basic rice productivity, which is the major crop in both states. In short, there can be no surplus generated in agriculture in Bihar with this productivity and large population to worker ratios. Agriculture remains a subsistence enterprise in Bihar.
What of the large range of non-agricultural sectors of the economy? For a comparative assessment we turn to the 2005 Economic Census of enterprises done by the CSO.
| 2005 Economic Census | AP | Bihar | ||
| Enterprise type | No. Enterprises | % of Total | No. Enterprises | % of Total |
| Animal husbandry, forestry, fishing | 994,937 | 24.39 | 34,974 | 2.86 |
| Manufacture and repair | 735,594 | 18.03 | 227,869 | 18.61 |
| Wholesale and retail trade | 1,259,022 | 30.86 | 700,451 | 57.19 |
| Hotels and restaurants | 140,646 | 3.45 | 52,623 | 4.3 |
| Public admin, education, health | 570,123 | 13.97 | 155,814 | 12.72 |
| Others | 379365 | 9.3 | 52,977 | 4.33 |
| Total | 4,079,687 | 100 | 1,224,708 | 100 |
The first and most striking feature of Table 7 is the fact that with a much smaller population, there are a staggering 3.33 times as many enterprises in AP (4 million) as there are in Bihar (1.2 million). This means that the population of AP is far more extensively served by all kinds of products and services than that of Bihar. A notable difference is the number of enterprises in public administration, health and education. The spread of these services in AP (5.7 lakhs) is vast compared to Bihar (1.5 lakhs). Perhaps this is one of the reasons why Bihar lags so far behind in human development indicators. Public and other basic services simply are not sufficient to serve the population of the state.
The major difference between the economies of the two states is exemplified by the first sector: animal husbandry, forestry and fishing. While in AP this sector has 24 percent of total enterprises, in Bihar they amount to only about 3 percent. For a land of many rivers and ponds, this is very surprising. But it reflects the deep distress in agriculture already seen in the productivity figures and in the 2001 employment profile. Another sign of this distress can be seen in the figures for the wholesale and retail trade sector. In AP this forms about 31 percent of all enterprises, but in Bihar it is the dominant sector forming a huge 57 percent of all enterprises. The distress in agriculture is driving people into putting up petty trade shops, rather than keeping a buffalo or some goats to sell milk and meat.
In fact, the sheer size of the AP non-agricultural enterprise in 2005, in comparison with Bihar, is underpinned by those productivity figures in agriculture. Without substantial surplus being generated by crop husbandry, it is not possible for there to be enough buying capacity in the population to support such a large and varied non-agrarian enterprise.
What is the hiring capacity of these enterprises and how does it compare between the two states? Table 8 gives the data from the 2005 Economic Census.
| Enterprise sector | No. of hired workers | ||||||
| 0 - 5 | 06-Oct | Oct-30 | 30 - 100 | Above 100 | Total | % 0-5 | |
| In Total | |||||||
| ANDHRA PRADESH | No. of Enterprises | ||||||
| Animal husbandry, forestry, fishing | 984,595 | 5,501 | 2,061 | 1,803 | 977 | 994,937 | 98.96 |
| Manufacture and repair | 696,287 | 25,501 | 6,974 | 4,661 | 2,171 | 735,594 | 94.66 |
| Wholesale and retail trade | 1,230,073 | 15,392 | 6,985 | 3,621 | 2,951 | 1,259,022 | 97.7 |
| Hotels and restaurants | 133,462 | 4,926 | 1,390 | 513 | 355 | 140,646 | 94.89 |
| Public admin, education, health | 507,481 | 36,415 | 19,619 | 4,729 | 1,879 | 570,123 | 89.01 |
| Others | 359,304 | 11,129 | 5,609 | 1,867 | 1,456 | 379,365 | 94.71 |
| Total | 3,911,202 | 98,864 | 42,638 | 17,194 | 9,789 | 4,079,687 | 95.87 |
| % Total | 95.87 | 2.42 | 1.05 | 0.42 | 0.24 | 100 | 95.87 |
| BIHAR | |||||||
| Animal husbandry, forestry, fishing | 34,775 | 181 | 12 | 3 | 3 | 34,974 | 99.43 |
| Manufacture and repair | 224,850 | 2,228 | 342 | 328 | 121 | 227,869 | 98.68 |
| Wholesale and retail trade | 697,510 | 2,677 | 150 | 74 | 40 | 700,451 | 99.58 |
| Hotels and restaurants | 51,976 | 568 | 62 | 12 | 5 | 52,623 | 98.77 |
| Public admin, education, health | 142,994 | 9,825 | 2,040 | 735 | 220 | 155,814 | 91.77 |
| Others | 50,992 | 1,360 | 499 | 104 | 22 | 52,977 | 96.25 |
| Total | 1,203,097 | 16,839 | 3,105 | 1,256 | 411 | 1,224,708 | 98.25 |
| % Total | 98.24 | 1.37 | 0.25 | 0.1 | 0.03 | 100 | 98.24 |
Let us look at the manufacturing sector a little more closely in Table 9. Diversity in this sector is highly desirable, since it means that employment will be available in many different types of work, requiring a range of skills, and the sector as a whole need not suffer due to a slow down of any one type of manufacturing.
| Type of manufacturing | AP | BIHAR | ||||
| Number | % in total | Above 100 Hired | Number | % in total | Above 100 Hired | |
| Food, beverages | 187,404 | 25.48 | 405 | 109,047 | 47.86 | 17 |
| Tobacco products | 70,186 | 9.54 | 385 | 4,099 | 1.8 | 1 |
| Textiles | 121,393 | 16.5 | 303 | 8,263 | 3.63 | 7 |
| Apparel, dress, fur | 124,104 | 16.87 | 318 | 17,685 | 7.76 | 0 |
| Wood, cork, straw products | 68,457 | 9.31 | 98 | 15,561 | 6.83 | 1 |
| Brick, glass, ceramic | 29,702 | 4.04 | 87 | 14,810 | 6.5 | 77 |
| Metal products | 15,112 | 2.05 | 59 | 7,621 | 3.34 | 3 |
| Machinery, equipment | 11,440 | 1.56 | 41 | 5,512 | 2.42 | 0 |
| Furniture | 68,143 | 9.26 | 111 | 26,627 | 11.69 | 3 |
| Others | 39,653 | 5.39 | 364 | 18,644 | 8.18 | 12 |
| Total | 735,594 | 100 | 2171 | 227,869 | 100 | 121 |
There are lessons here for Bihar. A much more diverse manufacturing sector in AP is seen to be relying less on food and beverages (which includes agricultural produce processing) and furniture than Bihar. Given the hiring capacity of these enterprises in AP, many of the units in textiles and apparel are larger ones. These are areas in which Bihar has traditionally had plenty of skill, with handloom and khadi in widespread use. However, this entire sector has been allowed to decline considerably, leading to the disproportionate dominance of food and beverage manufacture. In fact, in Bihar, the largest employers, hiring more than 100 workers (77 out of 121 hiring 100 or more workers in Table 9), are brick making units. There is no other enterprise worth mentioning.
Boosting agricultural productivity in major crops and diversifying its manufacturing sector, especially based on available skill within its population is the way forward for Bihar. The first is imperative. Without a substantial increase in agricultural productivity, its peripherals such as animal husbandry and fishing will not take root, creating substantial employment in the primary sector as a whole. In manufacturing, it will boost agricultural produce processing units. It is also time Bihar started taking note of the vast skill pool available within its borders, to target investment in key areas, such as textiles, to diversify away from brick making.
There was considerable dismay among the MLA’s at the sheer distance between Bihar and AP. Some of these MLAs had visited AP and other states as part of their duty, and were aware that it was, by and large, a more advanced state. However, the scale of that advance and the work that was needed to catch up was overwhelming. For a while, there was very little discussion, and some MLAs simply stopped thinking at this point. However, a few persevered and made a spirited attempt to identify areas in which they could move forward, especially for the specific case of Madhubani. One of them made the effort to get me a hearing from the State Deputy Chief Minister, who was also the Finance Minister. A report of all the work done was given to him. He appreciated the effort.
One such area was sugarcane cultivation and sugar production. Madhubani had 3 sugar factories until about the mid-1990’s. They were government owned and there was large scale sugarcane cultivation. MLAs had personal memories of farmers living very well by cultivating sugarcane and selling it to the local sugar mill. However, all mills had shut down, mostly due to poor labour management and corruption. Now, there was no sugarcane cultivation in the district. There had been some recent attempts to revive the sugar mills, but these too had failed due to procedural and administrative delays and inefficiencies.
The other area was spinning, weaving and production of ready made clothing. Again, there was a great deal of local expertise and skill available and these could be ear marked as priority areas.
At this point in time, assembly elections for Bihar were announced and within one day, everybody lost interest in the whole exercise and all conversations stopped. They were apologetic about this, but were firm in their assessment that their ability to give time and energy for this was almost zero, as they now had to engage in the whole business of getting a ticket and if successful, campaigning for the elections, for themselves and others.
The few conversations with senior administrators were as expected. They knew all this, and much more. But this was a very large problem, and nobody at the local level could be expected to engage with it or have any input for its progress in any useful way. The State would have to take care of it. And it was doing so.
Can data on the economy be opened to local decision-makers? There is a vast amount of highly detailed data available for district economic planning. Most of it is never used. It can be turned into intuitive insights by using maps to create awareness and discussion about the details of the economy of the district, its evolution over time, and its regional imbalances. It is also possible to give very targeted, on demand, comparative analysis to push the discussion into creative and feasible ideas for paths of development. Local elected representatives are most suitable for this task, as they are quick to point at the reasons behind the trends, and come up with solutions based on their deep knowledge of local conditions and possibilities.
However, for this to become a vital process that produces results on the ground, some critical links in governance need to be fixed:
The presence of at least one agency, preferably from within the government, to take charge of scouting for data that can be used at sub-district level, and to analyse it in consultation with local elected representatives, from village to State. An agency that is tasked with listening to the concerns inherent within the decision-making process and which responds to them by creating data based inputs to widen the arena of possibilities. This can be a combination of the planning department and the National Informatics Centre (NIC) offices at the district level. While the former participates in the planning process, the latter has the technological know-how to come up with imaginative and locally adapted data based inputs. In principle, they already have these mandates. However, they need to be far more proactive in implementing them. The process needs to be continuous and highly interactive.
Union and State agencies that collect data at district and sub-district level need to be proactive about advertising its presence, to begin with, and its usefulness. This can be done through the offices of the district statistical officers, who are tasked with collecting the data in the first place.
CSOs can also help by pointing at different sources of data, the many issues on which it might contain insights and methods to extract and present these.
The presently non-existent link between the village and district representatives and those at the State level needs to be established and strengthened. They need to work to an agreed plan and then mobilize politically at the state level to push it through. Depending on non-governmental actors for local planning has reduced the possibility of such planned political economic governance further. While a great deal of attention is being paid to creating awareness and capacity for local governance, no effort is being made to integrate the MLAs into this process by giving all levels a clear idea of their roles. It has created confusion, resentment, turf wars and false competitions, with the casualty being effective local governance.
In fact, the progress of decentralisation depends a great deal on a dynamic and energised engagement between local and State level representatives. This will clarify many issues of devolution of functions to local governments on the one hand, and their inputs to State level policy making on the other.
CSOs need to work with the political governance process and help to bring the various levels together on specific issues for better results. Using the bureaucracy as the only ally is highly limiting and ultimately unrealistic. They need to listen to the concerns of all parties and present acceptable solutions for striking local compromises. While they have done a great deal to create empowerment of the local governance process, they have not been able to mediate a creative political dialogue between levels of government, rather than pitching them as adversaries.
These suggestions are by no means simple to implement. In fact, political will is required both within and outside government to make some of them workable. Institutional egos will have to be set aside and political power sharing will have to be negotiated on a scale that at present looks very difficult. However, if one looks at the distance the country has travelled along the decentralisation path, there can be a great deal of hope of achieving these. Pressure created from below, based on an understanding of expanded notions of local governance and its place in regional development, is going to be the crucial factor. And open data is likely to play a pivotal role in building that pressure, to make it focused, and to create realistic bases for working together across political divides. There is much to celebrate in the sheer depth and diversity of India’s data regime and its accessibility. But a long way to go before the fruits of these labours become the bedrock on which governance rests.
This work was funded by The Asia Foundation and their continued interest is gratefully acknowledged. Lalitha Kamath, Asha Ghosh and Nick Langton were supportive of the unexpected turns that the work sometimes took, and their accommodation of these changes resulted in some very rich insights in governance. Shankar K.N. of Latticebridge Infotech gave enthusiastic support long after the actual contract was completed. In field work, the help and support provided in Madhubani by Parmeshwar Singh Negi was invaluable. Friends at the Bihar Seva Samiti showed Madhubani in all its beauty and despair, Anaro Devi provided an insiders’ view of the travails of her district, and these remain memorable lessons. The MLAs of Madhubani were gracious in their time in the face of persistent requests for discussions and finally came up with ideas that if taken forward in the future will surely benefit the district. And finally, editors of this special issue, who with great patience helped shape a field report into a detailed commentary on the place of data in planning, are gratefully acknowledged.
[1] Glover Write, Pranesh Pakash Sunil Abraham, Nishant Shah (2010). Open Government Data Study: India, http://www.transparency-initiative.org/wp-content/uploads/2011/05/open_data_india_final.pdf
[2] The Constitution (Sevety-third Amendment) Act, 1992, http://indiacode.nic.in/coiweb/amend/amend73.htm, The Constitution (Seventy-fourth Amendment) Act, 1992, http://indiacode.nic.in/coiweb/amend/amend74.htm
[3] As one State legislator in Maharashtra put it: Am I supposed to take orders from the GP president now?
[4] Manual for Integrated District Planning, Planning Commission, Government of India, 2008, http://planningcommission.nic.in/reports/genrep/mlp_idpe.pdf
[5] Status Report: PRI Capacity Building and Training in India (2011), http://www.pri-resources.in/MaterialUpload/Status Report 2011.pdf
[6] As one ZP president told this author: The administrator does not have the passion that I have, to push hard for what we want. They are afraid of their seniors at the State level and never argue with them.
[7] Economic Census data for 1998 was provided by the Indian School for Political Economy, Pune, India.
[8] There are manipulations of records by officials to project better (or worse!) performance in education and health, for example. Data on school attendance and maternal mortality is seen to be most prone to these. NGOs, press, local activists and politicians are some of the contestants. However, their success rate in getting such data corrected is difficult to track. Many district statistical officers, the nodal office between the district and the State, complain that they get ‘bad’ data from different departments, but can do nothing about it.
[9] National Informatics Centre, Government of India, http://www.nic.in/home
[10] Bihar Infrastructure Mapping – Geomatics Oriented Application Model, National Informatics Centre, Bihar, http://gis.bih.nic.in/Map/PlanningAtlas.aspx
[11] There have been instances in the author’s observation where district level administrators have colluded with State level planners to undercut a ZP plan to accommodate higher level political functionaries.
[12] Wolfgang Hoeschele (2000). Geographic Information Engineering and Social Ground Truth in Attappadi, Kerala State, Annals of the Association of American Geographers, Vol. 90, No. 2 (Jun., 2000), pp.293-321
[13] Madhubani District Vision 2020, Submitted to District Planning Committee, Madhubani, Bihar, by PRIA, http://madhubani.bih.nic.in/
[14] Census maps for Madhubani district were digitised by Latticebridge Infotech Pvt Ltd.
[15] Source: Directorate of Economics and Statistics, Department of Agriculture and Cooperation, Ministry of Agriculture, Government of India