[Note: Work in progress: not for quotation or attribution without the written permission from the author.]
Success and Failure in Rural Development:
A Comparison of Maharashtra, Bihar and Bangladesh

Harry Blair

Political Science Dept., Bucknell University
E-mail: hblair@bucknell.edu
Peasant Symposium Draft copy, 10 May 1997

I. The Political Economy of Rual Development in Bihar and Bangladesh
II. Rural Development in Western Maharashtra
III. Explanations: Why are the Two Regions So Different?
*History *Demography *Social Relations *Political Culture
IV. Implications for Rural Development Policy
Table 1. Total Population and Largest Cities, 1981/1991
Table 2. Literacy in Maharashtra and Bengal/Bihar as a percentage of the population, 1931-1991
Table 3. Principal Ethnic Groups in Bihar, Maharashtra and Bangladesh
Figure 1. Political-Economic Development Linkages in Bihar & Bangladesh, Mid-1980s
Figure 2. Political-Economic Development Linkages in rural Western Maharashtra: Ahmednagar District, Mid-1980s
Figure 3. Polilitical-Economic Development Linkages in Bihar, Mid-1990s

Apologia pro exitu suo

This paper is offered to the Hauserfest as an effort to promote and provoke dialogue on what I consider to be Walter's central lifetime research project: what makes rural Bihar the fascinatingly and exasperatingly backward place that is? I originally wrote the paper in the spring of 1988 for an AAS conference organized by John Echeverri-Gent, conceiving it as the outline of a book project I then wished to take up that would compare rural development in the three places I had studied at some first-hand length: Bihar (since the mid-1960s), Bangladesh (since the early 1970s) and Maharashtra (beginning in the early 1980s). They offered a nice set of contrasts -- the relatively successful track record in Maharashtra, particularly the sugar-growing area in the western part of the state, as against the abysmal failures experienced in Bihar and the mostly (though not completely) bleak trajectory established by Bangladesh. In the paper, I believe I made a good start in sorting out the most salient aspects of the history, demography, social relations and political culture that account for the differences between these three regions. I looked forward to exploring these patterns in my future work.

After writing the paper, however, my own research trajectory changed markedly. Instead of undertaking the (undoubtedly over-ambitious) broad-gauge comparative study contemplated in my AAS paper, I wound up putting the paper aside unpublished and narrowing my focus considerably to Bangladesh for several years. Then in the early 1990s, I moved off in a different direction altogether, engaging myself in the global democratization support initiative then being undertaken by the international donor community. The experience has been an exciting one, but a part of the opportunity cost has been an almost total neglect of the subcontinental research that attracted me for so long. The 1997 Hauserfest offers an excellent opportunity to rekindle that interest, and to assess whether the distinctions that seemed so important in 1988 still do almost a decade later.

A quick glance at the three regions finds the political climate considerably changed, to say the least. After four decades and more under uninterrupted Congress rule, Maharashtra has taken a lurch to the Hindu right with its Bharatiya Janata Party/Shiv Sena government. In Bihar the rise of the backwards that didn't quite get established in the Karpoori Thakur era has now achieved an ascendancy in Laloo Prasad Yadav's ministry. And in Bangladesh, the Ershad dictatorship of the 1980s has given way to a parliamentary democracy. At other levels, though, things do not seem so different. Western Maharashtra's sugar belt continues (I think) its relative prosperity. In Bihar the Rs 950 crore fodder scam currently swirling around Laloo Prasad Yadav seems strikingly reminiscent of the huge fertilizer scandal that beleaguered Chief Minister Jagannath Mishra in the 1980s, both in its scope and in the strong probability that Laloo's political fortunes will be not be harmed any more by the scandal itself than Jagannath Mishra's were in his time. And in Bangladesh, the principal defect in the political system -- the inability of the parties to accept even the most basic operating rules of comity -- continues to threaten fundamentally the basic stability of the polity.

I should also mention that the foreign assistance picture has changed quite radically since the mid-1980s. At that time budgets were relatively large, and policy expediency in pursuit of Cold War objectives often gave dictators precedence over democracy in donor calculations. Today budgets have become much reduced, while support for democracy has attained a priority close (and maybe even equivalent) to that accorded to markets. It is difficult, in a word, to imagine foreign aid any longer assuming the role it played in Figure 1.

But that is all by the way. The main issue that I hope the paper will lead the Hauserfesters to consider is this: does it still make sense (assuming that it did at least to some extent in 1988) to look at these three areas of the subcontinent in the ways I have suggested, or has the situation changed enough to demand new analytical lenses and perspectives? And if so, what new approaches to comparative analysis can be offered?

I am very much looking forward to our gathering later on in May. See you all then.


Success and Failure in Rural Development:
A Comparison of Maharashtra, Bihar and Bangladesh

Harry Blair

Political Science Dept., Bucknell University
E-mail: hblair@bucknell.edu
Draft copy, 10 May 1997

This paper represents an outline of thoughts that have taken about fifteen years to put together on rural development in several parts of the Indian subcontinent.(1) It is thus more the skeleton of a book-length analysis than a complete presentation in itself. I offer the outline in hopes of receiving comment and criticism that will be useful in turning it into the much longer treatment that the topic deserves.

What I am basically trying to do here is rough out a set of ideas as to why rural development has been more or less successful in Western India, while it has unquestionably been as yet a failure in most of the eastern part of the subcontinent. Specifically, I wish to concentrate on Maharashtra and within it on the area known as Western Maharashtra as a relative success story, though the concepts to be sketched out here in many ways apply to Maharashtra as a whole and within broad limits to Gujarat as well. As cases of failure in rural development my focus will be on Bihar and Bangladesh, though much of the reasoning would apply with equal force to Eastern Uttar Pradesh and Madhya Pradesh (and in very broad terms West Bengal fits in too, though its rather different history in recent decades has put a distinct spin on things there, so much so that I will not include it in the present analysis). Between the two cases of failed development, Bihar is clearly the more intractable, though perhaps not absolutely and hopelessly so, while in Bangladesh the scope for silver linings to the dark clouds of reality is somewhat greater.

The object of the exercise here is partly to seek and build an understanding of the wider patterns of development in rural South Asia, as any worthwhile effort in social science should do, but my motives also go beyond that level to search for policy options. To be blunt about the enterprise, I wish to ask: are there any lessons in the experience of Maharashtra that would be valuable in formulating rural development policy for the eastern portion of the subcontinent? And to anticipate my answer, I believe there are some answers in the Maharashtra experience, albeit tentative ones and ones that will at best take a fairly long time (at least as policy makers think of time) to bear fruit. There are indeed some useful things to be learned.

At this point, I should lay out my credentials and lack of them for the task at hand. I first tried to fathom Bihar in my dissertation research during 1966-67, in an election study that was essentially political anthropology. Not until the early 1970s did I become concerned with rural development, when I had a chance to do a field study in Bangladesh. Since then I have returned to both Bihar and Bangladesh a number of times, and while I do not know either with the grasp and depth of specialists like Walter Hauser and Pradhan Prasad on Bihar, or Peter Bertocci and Rounaq Jahan on Bangladesh, I am for the most part comfortable with the two regions. In working on rural Maharashtra, however, I am a relative tyro, having been there the first time only in 1982 and that merely for a couple of months. Two more recent visits have helped, as well as a good deal of reading, but I remain clearly an academic arriviste in the Deccan of Western India. Still, I hope that I have picked up enough to gain some insights into what has been going on there and why things happen as well.

And last, before getting on with things, a few definitions and terms are in order. "Rural development" is used here in a fairly wide sense; it encompasses economic growth and distributional equity (across lines of class, ethnicity and sex) in the countryside, as well as advances in such "life quality" measures as literacy, infant mortality and life expectancy. Political and institutional development are also a part of rural development, but more as means than as ends. In particular, popular participation in decision-making, which some would see as a worthwhile end in itself (a view with which I would not disagree), is treated here as instrumental to such ends as better public health or social welfare. As a shorthand, I shall refer to Western Maharashtra as WM, and to Bihar/Bangladesh as BB.

I. The Political Economy of Rural Development in Bihar and Bangladesh

The dynamic presented in Figure 1 is a depressingly familiar one to anyone who has worked in the eastern part of the subcontinent, whether as an academic researcher or a rural development practitioner. Its only redeeming virtue is an elegant simplicity which makes it dismayingly easy to lay out in a few words. There are three groups of significance here in this semifeudal(2) rural political economy, connected together by an age-old binding of patronage, support and exploitation. The first level is the state , which lies uneasily at the apex of the structure, concerned principally to maintain itself in power.(3) Its primary objective with respect to the countryside is to assure a relative peace and to prevent serious disorder from breaking out (previously, an additional objective was to extract rents, but this revenue function has attenuated virtually to the vanishing point in recent decades). The state attains this goal by forming and keeping a de facto alliance with village-level landowning elites, whereby government guarantees property rights (despite periodic rhetorical gasconades fired off promising land reform) and provides patronage in the form of rural development program spending, in return for which rural elites keep local order and support the regime.

Village elites in their turn exercise a control over the lower orders through the customary machineries of patron-client relationships, money-lending, sharecropping, wage labor, dominance over institutions of local self-government, and of course the traditional goondas and lathials kept on retainer to enforce elite authority. The objects of these attentions are of course the remainder of the village population, the small and marginal farmers, the artisan families and those who are without either land or traditional craft skills, i.e., the landless agricultural workers and their families, who form the residual occupational category for those who cannot find employment elsewhere. In return for the subsistence that is provided to them and in consequence of the control that is exercised over them, these village poor and landless work for their patrons and provide support to them by accepting the status quo in the countryside. At times, things wear too thin and protest breaks out of a magnitude that cannot be easily subdued by the village thugs; on these occasions, the state sends in sufficient armed constabulary to subdue whatever trouble has arisen. But such outbreaks are relatively rare and isolated, such that the usual systems of control generally suffice to keep things in order.

Whatever resources come into this system, whether from foreign donors or the central government's development budget, fit easily into it and even reinforce it. Thus projects intended to provide agricultural credit or furnish irrigation equipment or set up small-scale rural industries become perverted to funnel funds into the hands of local elites. But this pattern of rural development perversion should not, as Keith Griffin among others has pointed out, be viewed as "failure", but rather should be seen as evidence of a successful attempt to maintain the countryside as it is. The overall rural economy, then, is an impressively homeostatic structure, into which an independence movement can come, political parties can flourish (even competitively for much of the time), a zamindari system can be abolished, sizeable infusions of development funding can be poured, hordes of development administration personnel can be deployed, and a "Green Revolution" in agricultural technology can be introduced, all without producing more than ephemeral change.

That all these developments have served primarily to reinforce an enduring reality can be explained in large part by the overall orientation of the landed class, which is essentially to preserve its position, to maintain control over the lower orders, rather than to maximize income or wealth. For their part, the rural poor find that their most pressing need in an environment of steadily increasing pressure on the land and surplus labor eking out a living from that land is simple survival. If acceptance of elite dominance is the price to be paid for survival, then the bargain is not a bad one, and in any event, the alternative is most likely not to survive.

The picture presented here is to be sure a greatly simplified one. There are other classes or strata in the countryside beside elites and the poor. One significant group here is a "middle farmer" segment of the peasantry that is attracted by the prospect of increasing income through raising productivity, thereby offering a way out of the semi-feudal economy. This is the heart of the middle caste Lok Dal constituency that a decade ago seemed on the verge of consolidating a breakthrough in Bihar, but which has since succumbed to a restoration of high caste, large farmer control (in Bangladesh there has not as yet appeared any political party to champion the interests of this stratum, though the "progressive farmer" spirit certainly does exist to some extent). And there has been in recent years a pattern of increasing unrest at the bottom of the class/caste spectrum, particularly in Bihar, which has seen a growing activism among landless Harijans, along with the most brutal repression of their cheekiness at the hands of rural elites and the police establishment acting in the interests of those elites. There is also some evidence, admittedly skimpy for the most part, of rural development efforts achieving a modest success. A new bridge across the Ganges at Patna, for instance, has connected a large swath of previously isolated North Bihar with the larger economy and is bringing a more rapid economic growth to the area. In Bangladesh, fertilizer use has grown substantially over the last fifteen years. Rural electrification has made some progress in both areas.

But these qualifications are comparatively minor, indicative perhaps of social transformations slowly and even glacially emerging, but as yet such changes are but small currents and isolated eddies set in a huge and essentially stagnant backwater. Plus ça change, plus c'est la même chose remains the basic all-purpose aphorism for the region.

II. Rural Development in Western Maharashtra

The institutions and linkages portrayed in Figure 2 apply to Ahmednagar District, which might be said to be the central or core region of the Maharashtrian Deccan, but mutatis mutandis the analysis here would apply to the remainder of Western Maharashtra as well and, in a somewhat modified fashion, to the state as a whole.(4)

The first thing to note is that the structures of Figure 1 are also present in Figure 2. Village level elites (here labeled "big and middle farmer," noted as A in the figure) and the rural poor and landless (B) are still there, as is the state government, though the latter is renamed "Representative political system (D)," for reasons which will become apparent as the discussion proceeds. But there are a good number of other institutions present as well. In a paper of this scope (which is already threatening to become too long), it will not be possible to examine each of these institutions, but I will endeavor to point out the salient features of the system.

What is important here is not only the institutions themselves, but also the linkages connecting them together. Both aspects are far more highly developed in WM than in BB. In general terms, the institutions of WM are also present in BB, but in the former case, they have become functionally specific, in the Parsonian sense, such that they can and do perform the tasks which they are nominally assigned, whereas in BB the rural political economy virtually ensures that whatever the formal purpose of an institution, its true function is simply to grease the machinery supporting the status quo. Thus development projects like the Mula irrigation scheme in WM (J in Figure 2) do in fact provide water to farmers in timely fashion, sugar cooperatives (E) have become immensely powerful engines of rural development, and the Employment Guarantee Scheme (K) has established a welfare "floor" which guarantees a wage to everyone in the rural areas of the state.

Several features of Figure 2 deserve mention, even in a treatment as brief as this one. The sugar cooperatives are an outstanding success story in the annals of rural development. Begun in the early 1950s, they now process the overwhelming majority of a steadily increasing sugar crop in the state, provide immense direct benefits to their members, and exercise a large (some would say inordinate) influence in the state as a whole. Various reasons can be cited for their success, perhaps chief among them the several peculiarities of sugar cane production which make it relatively easy for cooperatives to maintain a close linkage to their growers. To begin with, sugar is a crop that, unlike foodgrains, needs processing before it can be used by consumers. Some variations on that processing can be fairly rudimentary, such as gur (crude brown sugar) making, which consists essentially of squeezing the juice from the cane and boiling it down in large vats.(5) But the major demand these days is for refined white sugar which sells at a considerably higher price. Refining is a highly sophisticated technology, necessitating a very large outlay in terms of investment. The point here is that before sugar can be sold to consumers, it must be processed.

It is this need for processing that gives the cooperative its leverage, for growers find they have to sell (at least most of) their crop to the cooperative, and when they deliver it, the cooperative can deduct from the price the cost of whatever loans have been made. There is in other words a built-in mechanism for recovering loans, a feature which allows the coop to remain solvent, thereby avoiding the problems of loan default that have so devastated coops elsewhere. A second quality of sugar here is that it is highly perishable. After cutting, the cane begins immediately to lose its sucrose content; it "spoils as rapidly as milk," in the local aphorism. Accordingly, long storage to wait for a price rise or transport to distant processors who might offer a better price is just not feasible. And third, sugar is an extremely thirsty crop; its need for water is far in excess of any other crop grown in the region, which means in semi-arid WM that sugar is dependent on reliable irrigation.

But these are far from the only reasons for sugar's success. In addition to these factors of crop production and processing technology, there has been a very strong connection to the political sector (D in Figure 2) in the form of massive subsidies bestowed on the sugar cooperatives. These subsidies come in several forms. First, the state government has provided large-scale loans to sugar cooperatives to help them raise the sizeable start-up capital required. Second, the water supplied through the various irrigation schemes has been priced to farmers at a fraction of its true cost.(6) Third, electricity used to lift water for sugar has been heavily underwritten by the Maharashtra State Electricity Board.(7) And last, government has provided cane price supports that mean the domestic price of sugar in India has been a multiple of several times the world price.(8) Thus water has been a massively subsidized input provided by state at high cost, to be wastefully used in growing and processing a crop to be sold at a very handsome price and profit in lieu of what would be a rather cheap imported commodity.

The mechanism involved in securing this bounty to the cooperative sugar sector has been one that would be very familiar indeed to North Americans or Europeans. Simply put, the cooperatives have used a portion of their profits to back candidates for political office, and have then put pressure on those elected to obtain and maintain massive subsidies to sugar. The story sounds strikingly like the saga of Louisiana cane sugar and western beet sugar growers in the United States, who have also been able through political pressure to push the domestic price of sugar in the United States up to three or four times world levels.

The Employment Guarantee Scheme (K in Figure 2, known throughout the state by its English acronym, EGS) is another remarkable institution in Maharashtra. In effect for almost fifteen years, it virtually assures a daily job in unskilled labor at minimum wage to anyone in the rural areas of the state who asks for one, subject to the stipulations that there can be no unmet demand for agricultural labor in the area and that the wages paid by EGS may not be higher than those prevailing in the local agricultural sector. In the early 1980s, EGS provided annually about 150 million person-days of work, at a cost of about Rs 140 crores (US $115 million) a year. There are a number of problems with the program, to be sure. Much of the rural works constructed, for instance, turn out to be land improvements such as bunding and leveling on land belonging to richer rural households, rather than community assets. Second, in some areas (particularly the drought-prone districts of WM), officials are running out of projects that can be taken up through EGS: just about all the work that could be done has been done, leaving less and less for future years. Third, the EGS itself can be seen in a cynical light as a device perpetrated by urban elites (C in Figure 2) for keeping rural hordes in the countryside by bribing them to stay out of Bombay (it is funded mainly through a combination of urban professional taxes, sales taxes and general state revenues, most of which come from Bombay). And last, there is considerable political pressure on officials to employ the program to nurse their constituencies. But the long and short of it is that the EGS has built a rural social safety net that does not exist in like form anywhere else in the country.(9)

It is of course true that similar institutions exist in BB. There are cooperatives, representative political institutions, and rural works schemes in both regions. But in this more easterly environment, rather than to raise agricultural production, the entire cooperative structure serves mainly to channel state funds into the hands of rural elites by allowing and even encouraging them to take over cooperative units, steer the loan monies to themselves, and then default on their overdues. Institutions of local and state government are dominated by these same elites, who are thereby aiding and abetting themselves in perverting the cooperative structure, while doing nothing for the rural poor and landless. Likewise, programs specifically aimed at such groups, like rural works schemes, are also swept into the politics of patronage and corruption, yielding little but further rewards for the rural rich.

Needless to say, I have exaggerated the picture somewhat here. It would be hard to find many successful coops in BB, but there is some evidence that the political system is having to pay some attention to the middle class/caste farmers in Bihar that formed the backbone of the Lok Dal, if only by buying their allegiance to the Congress Party.(10) In Bangladesh, the Rural Works Programme that began under Ayub Khan in the early 1960s was a case study in all the abuses outlined above, but there are other efforts that have been better managed, such as the current CARE Food for Work program, which is currently generating some 20 million days of work annually in Bangladesh and is a model of development administration.(11)

But the overall tenor of the rural development experience in WM and BB could scarcely be more different. In WM there is some corruption and considerable public waste offset against a pattern of steady and significant rural development that has had a real effect in raising income and living standards across the entire spectrum.(12) In BB on the other hand, there is some indication of development here and there, perhaps, but the overall pattern has been one of stagnation and decay. The question, then, is why should things be so different in the two regions? It is to this issue that the discussion turns in the next section.

III. Explanations: Why Are the Two Regions So Different?

The argument so far has been reasonably straightforward. The details remain to be filled in and fleshed out, but the evidence for development in WM and stagnation in BB is so abundant that there should be relatively few difficulties in doing so. Explanations of these glaring differences are of course more problematic, but then good social science should explain, as well as portray. Moreover, as I mentioned in the opening paragraphs to this essay, I do have a policy recommendation agenda, and a good understanding of the reasons behind the differences between the two areas is a necessary prerequisite to such an endeavor.

As might be expected with anything so complex as success and failure in rural development over the past four decades, possible explanations abound. For convenience, they are grouped here in terms of history, demography, socio-economic and cultural dimensions, and political style, an ordering which in a very loose sense constitutes a spectrum between what is more causal and what is more consequential.


1. Land tenure systems. The basic outlines of the differences between the ryotwari and zamindari systems of land revenue and tenure are sufficiently well known that I need not spend time on them here. Their consequences are also well appreciated by those who have worked in the vineyard of the subcontinent's rural economy. In many, many ways the former zamindari areas of India and Bangladesh still labor under the historical weight of a system that discouraged husbandry and land improvement among the peasantry while rewarding indolence and parasitism among the landlord class. On the other hand the relatively freehold tenure enjoyed by the peasantry of the Bombay Presidency in the British period gave it a head start as a class producing independently for a market. For the former, the feudal/semifeudal concerns of status and control were paramount, while for the latter, landowners were more concerned with maximizing income.

2. Urban-rural linkages. The mofussil of the Bengal Presidency constituted a hinterland supporting with its rents a cosmopolitan elite in Calcutta, whereas the rural areas of Bombay Presidency, though in many ways a hinterland to Bombay city, did have a cultural self-sufficiency and self-identity with centers of culture and economic dynamism like Pune, Ahmedabad and Vadodara that were lacking in the Bengal hinterland, where Dhaka, Patna, Cuttack, etc., were just not on a plane with the western towns. When the various segments of territory from the Bombay Presidency, Central Provinces, Hyderabad and a wide scattering of smaller princely states all joined together in the states reorganization of the 1950s to make up the Maharashtra state that emerged in 1960, Nagpur and Aurangabad replaced the Ahmedabad and Vadodara that disappeared with the creation of Gujarat state. In sum, Maharashtra developed a relatively integrated regional culture and economy.

In BB, on the other hand, an urban regional infrastructure did not develop after the 1912 splitting off of Bihar and Orissa. The new province continued to be dominated by Calcutta down through the partition in 1947. Even today, Bihar has nothing like a developed urban regional economy, as is clear in Table 1, where it will be observed that this state of almost 70 million cannot support a single city (Patna) with over one million in size. And in fact the cities ranked second (Dhanbad), third (Jamshedpur) and fourth (Ranchi) all are industrial towns connected primarily to Calcutta and having little to do with the rest of Bihar outside the Chotanagpur industrial belt. Thus in terms of the socio-economic-political-cultural entity that is Gangetic Bihar, an area of some 56 million people in 1981, the second largest municipality was Gaya, which at that time had fewer than 250,000 souls. For Bangladesh, the story was one of even less urbanization, as East Pakistan became cut off from its Calcutta metropolis in 1947, leaving the new province of 40 million people with its capital Dhaka having a mere 200,000 inhabitants. Even by 1961 Dhaka still had less than 600,000 people, though its growth as the capital of independent Bangladesh has been quite rapid indeed, so that by 1981 it boasted almost 3.5 million people. The drop-off in the rank-size listing in Table 1 remains remarkable, however. After the first two cities Dhaka and Chittagong), number three (Khulna) has only 620,000 and the fourth highest is less than 200,000. In other words, Dhaka gives some (albeit mixed) evidence of becoming to Bangladesh what Calcutta had earlier been to greater Bengal -- a parasitic city that both drains the countryside and in the process prevents other cities from becoming more than overgrown villages.

In these respective processes of urban growth and non-growth, the linkages that tie city to countryside progressed quite differently in the two areas. Calcuttta had (and to some extent Patna still has, though by now in Dhaka the situation is rather different, I believe) an essentially exploitative relationship to its mofussil, extracting rural rents to maintain a sophisticated urban lifestyle. Bombay, Nagpur and Pune are scarcely innocent of exploiting their hinterlands, but in the bargain they provide an array of developmental assistance (state services, finance, trade, etc.) that has actually promoted economic growth in the countryside.

3. Martial history. WM was the base of Shivaji's operations in the 17th century, when the Maratha Confederacy began to push out the Mughal Empire and then gave a fair challenge to the East India Company for dominance of western India. There is thus a legacy of martial success and self-assertion in WM that contrasts markedly with the experience of BB, which has been under the sway of "foreign" (though some of them indigenized themselves quite well) rulers for the last thousand years or so. One obvious manifestation of this is the equestrian statues of Shivaji that one sees all over WM (almost always in a traffic circle fenced in by wickets shaped like military shields with crossed spears), proclaiming the martial legacy of the citizenry. Self-assertion vs subjugation by foreigners is the theme here.


The social makeup of WM has one clearly dominant caste, while in Bihar there are several large and powerful caste groups all contending for dominance. The Marathas constitute about 35% of total population in Maharashtra (people commonly say 40%, but as best as I can tell from the 1931 census(13) it seems to be about 35%, assuming population growth pari passu among all the state's communities). The next largest group are the Mahars, a Scheduled Caste community with about 10% of the population. They do have some political clout (the Nava Buddhist movement and the Dalit Panthers, for instance, are mainly Mahar in their leadership and composition, I think), but they are not in any serious way rivals to the Marathas for dominance. The result is that Marathas determine the political leadership of the state. They do have factions among themselves, to be sure, but this is not enough to allow other groups to move into control. Even when a non-Maratha has been in charge, e.g., V. P. Naik as chief minister in the 1960s, his sitting on the gaddi was possible only because a substantial segment of Marathas agreed to it.

In Bihar the largest Hindu community has about 11% of total population (again making the same assumptions on growth of different groups at the same pace since the 1931 census) -- the middle ("backward" in the Bihar lexicon) caste Yadavas. Then there are three upper caste ("forward") communities in contention (Brahmans at 5%, Bhumihars at 3% and Rajputs at 4%), as well as two other "backward" caste groups (Kurmis and Koiris, both at 4%). The consequence in Bihar is that since the first Congress government in the late 1930s there has been intense caste rivalry over gaining control of the state's politics. It is not indulging in hyperbole to say that an incredible amount of time and energy has been and continues to be squandered on fights for caste dominance in the state's political system, leaving very little of either available for leading serious development efforts.

Bangladesh of course is 90% Muslim and so caste conflict is really not an issue there. The implication is an interesting one. Insofar as Bihar and Bangladesh are similar cases, and given that caste conflict is not present in the latter, then that same caste conflict cannot explain more than a part of the developmental backwardness of the former. Obviously, of course, the two are not the same and so caste can carry some of the explanatory freight for one of them, but the similarities in most other respects of their histories, cultures and economies are so striking that the explanations of their backwardness must be similar in many ways as well.

Social Relations

1. Interpersonal relations are generally less harsh in WM than in BB. In a palpable sense, people are nicer to each other. This manifests itself in all sorts of ways. For one thing, the superior/subordinate relationship in BB is characterized by arbitrary arrogance from above and a fawning sycophancy from below--the "ji huzoor" mentality pervades all hierarchical organizations. While WM is scarcely egalitarian, the relations between higher and lower status levels seems significantly less rigid and exploitative. Secondly, social behavior is more public-regarding. People in BB hardly ever stand in queues in public places like bus stops and post offices, they have virtually no civic consciousness in throwing out refuse, they drive and park vehicles with no regard to convenience of pedestrians.(14) In WM people are much more socially conscious on all these counts.

2. The status of women is much higher in WM. Women are more conspicuous in public and less restricted in habits of dress and movement. The whole milieu of purdah, which is so pervasive in BB is hard to discern in WM. In Pune women wearing dresses are so common as to be ordinary, while in Bihar and Bangladesh dresses are only for small girls, below the age of 10 or so. In WM women riding bicycles or motorscooters by themselves are piloting perhaps 15-20 percent of the two-wheeled vehicles on the roads, whereas in BB they are almost never seen doing so. Women are also better represented in the professions in WM--as doctors, professors, lawyers, etc. "Eve teasing" (the public abuse of adolescent females by males of the same age, including verbal harassment, pinching, punching, etc., but very rarely rape) is a growing pandemic in north India. In WM, on the other hand, it seems exceedingly rare. It is just not part of the culture there.

Perhaps the most telling statistics are those for female literacy, as can be seen in Table 2. The effort to educate females began in modern times in Maharashtra with Mahatma Phule (see below), but in terms of literacy things geared up only after the partition. By 1971 literacy among females in Maharashtra was getting close to what it was among males in Bihar and Bangladesh. Ten years later female literacy in Maharashtra had pulled significantly ahead of the male figures in Bangladesh (which had even retrogressed somewhat in the 1970s) and was just about even with Bihar in this respect.

3. There is a tradition of social reform movements in WM, stretching back into the 19th century if not earlier. Mahatma Jyotirao Phule was the first of the great social reformers, responsible in large part for displacing the small Brahman community from their traditional dominance in the Bombay Presidency and for beginning efforts at uplift of women. Dr. B. R. Ambedkar, the great Untouchable leader, is undoubtedly the premier social reformer of the 20th century (today, three decades after his death), and his statue is probably second or third in popularity in public places. But there are many lesser leaders as well. One way to put this tradition in perspective is to note that it is considered a worthy and respectable avocation for successful men in their later years to involve themselves in social uplift activities (or perhaps more important, there seems some social pressure on them to appear to be doing so, even if in fact they are not).

By contrast, the leading social reformer in Bihar up until quite recently has been Sachchidanand Sinha, whose great contribution was to lead the movement for a Bihar Province separate from the Bengal Presidency. This was "social reform" in a sense, but it was also very much and mainly an effort to set up a Bihar for his caste brethren the Kayasthas to administer and manage, independent of the "bhadrolok" Bengalis who had all the good jobs in running the Bengal Presidency. Similarly, the great hero of modern times in Bangladesh is A. K. Fazlul Huq, the "Sher-e-Bangla" ("Tiger of Bengal"), whose main contribution as perceived today was to establish the political dominance of the Muslim majority in Bengal during the brief popular governments of the late 1930s, thereby displacing Hindu control.(15)

In more recent times, of course, there has been Jayaprakash Narayan in Bihar, whose credentials as a genuine social reformer are very high indeed, while on the other hand Y. B. Chavan made his early career as a politician by masterminding the bifurcation of Bombay state into present-day Maharashtra and Gujarat, thereby establishing a state where Marathas could manage things (acting in the name of Maharashtrians generally, of course). But these are exceptions, I would argue. The impulse to social reform is not unknown in Bihar, but it is certainly rare. Political opportunism is definitely found in Maharashtra, and it is more than rare, but it is not so overpoweringly the motif that one finds in BB.

Political culture

1. There is a kind of accommodational approach in the politics of Maharashtra that is absent in the more vicious "winner take all" politics of BB. In Maharashtra political winners are expected to reserve some share of rewards for losers, or at least to keep the losers working within the system. Winners in Congress Party factional struggles do not try to purge losers from all rewards of political life, for today's losers are expected to stay in the game, maybe even to win later on. In my time there, in the second half of 1985, the main political issue of the time centered on Sharad Pawar, a former chief minister who had left the Congress to form his own party, a group which had lost badly in the parliamentary and assembly polls earlier. But instead of trying to eliminate him completely from political life (which is what the winners would have done in Bihar), the winners in Maharashtra were dickering with him on the terms of his readmission to the Congress Party (a move that eventually occurred in late 1987). This mellowness no doubt reflects the Congress' leadership's assessment of Pawar's own real strength, but what is interesting here is the mentality that says "let's make a deal" rather than "let's liquidate the other side while we can."

Another example is rural violence. This is hard to measure in any valid way, and knowledgeable people in both WM and Bihar assert that it has gotten much worse in recent years, to the point that the socio-political system is under great strain, perhaps even to the point of siege.(16) Still, the incidence of Harijan atrocities, goonda terrorism and the like seem to be much greater in rural Bihar than in Maharashtra. My impression is that communal violence is also a good deal less severe in Maharashtra, but I really don't know enough about this to be sure.

2. The standards of judgment for officeholders. The ways in which the culture evaluates political leadership is quite different in the two regions. In Maharashtra the principal criterion seems to be performance, with integrity in a strong second place. It is expected that there will be some corruption, leakage, lobbying for special interests, nursing of constituencies, etc., but it is also expected that government must deliver something concrete in promoting rural development (and development in general, for that matter). Programs intended to raise agricultural production must deliver inputs so that they get used to increase irrigation, get fertilizer onto the fields, make credit available to those who will use it to increase crop production, and so on. In WM's archetypal example, the Employment Guarantee Scheme must in fact provide unemployment relief for the rural poor (though it will also be beneficial to local politicians in the process, there will be some bungling and embezzling, etc.).

In BB patronage seems the main criterion for judging officeholders. It isn't really expected that inputs will ever be used as intended, that irrigation projects will actually deliver much water to farmers, that roads will be built to handle the traffic load they were designed for, etc. The first purpose of government development activity is to provide patronage and build allegiance for those who can deliver the patronage. A second and allied purpose is to maintain the rural gentry's control over the lower strata at village level by diverting the loot that comes down from higher level to reinforcing their own positions.

It probably goes too far to say that the local power structure in BB is hostile to any significant economic growth on the grounds that such growth might encourage the lower orders to be more demanding, for there are some (albeit feeble) signs of an agricultural transformation to a market-oriented agriculture going on. But there is certainly a massive reluctance on the part of the gentry to seeing any benefits go to the rural poor. And it is definitely a very long way from rural Bihar to rural WM in this regard. To sum up, one could ask: is an effective EGS even remotely conceivable in rural Bihar? To ask such a question, regrettably, is to know the answer.

A good illustration of the difference in standards of judgment of officeholders can be seen in the resignations of Chief Ministers A. R. Antulay and Shivajirao Patil-Nilangekar in Maharashtra. Both were forced out because of corruption scandals aired and publicized in the media -- Antulay in a cement allocation scam (awarding allocations of this perennially scarce commodity in return for contributions to a political fund) and Patil-Nilangekar in medical examination fixing scheme (getting the marks adjusted on the state gynecological exams so that his twice-failed daughter could pass). By contrast, at about the same time as Patil-Nilangekar was resigning in Maharashtra, a medical scandal was emerging in Bihar in which Chief Minister Bindeshwari Dubey was trying to "regularize" the appointments of (i.e., give permanent civil service status to) the one-third of the state's health service cadre who had been given their jobs on various ad hoc schemes over the previous eight years. Needless to say, more than a few of these appointments had been made on the basis of qualifications other than medical, but there was no threat to the chief minister on this ground.

Indeed, in Bihar it is inconceivable that a chief minister would be sacked on grounds of corruption or incompetence. There have been forced departures and even hints of scandal connected with them, but the ousters have always been mainly concerned with factional intrigues and power bases. The most notable example recently has been former chief minister Jagannath Mishra, who has been strongly implicated in massive corruption at the state fertilizer cooperative. He was dismissed by Prime Minister Rajiv Gandhi, but no one interpreted the departure as seriously related to the fertilizer scandal. Rather, it was because of factional shifts against Mishra in the Bihar Congress and because the personal dynamic between Rajiv Gandhi and the Mishra was highly negative (By contrast, Indira Gandhi, who had a well known "soft corner" for Mishra, kept him in office no matter what the aroma of misdeeds surrounding him). Perhaps the best way to put it is to observe that a chief minister in Bihar would be judged wanting if he were not fiddling with medical exams and cement allocations on behalf of relatives, caste-fellows and cronies.

4. The styles of politics. Politics in Maharashtra is characterized by what might be called a transactional style, in which people involve themselves in the political arena to exchange benefits. Trading and deal-making are the modes of participation. In BB, the dominant pattern is an allegiant style. One plays politics to build up and nurture a support base. Goods and services do change hands, but they do so as a means of buying and selling support, whereas in Maharashtra the exchanges are made with a view to using the goods and services to produce and build things. Of course there is a good deal of overlap between the two styles, for transactional politicians need their support bases too, and the goods and services traded by allegiance-oriented politicians do sometimes serve to increase economic and/or social productivity. But in the main, there is a meaningful distinction here. The sugar nexus in WM is illustrative of the transactional style, for all the vote-trading, surreptitious candidate funding and so on really does lead to more investment in rural infrastructure, while in BB, the horsetrading that goes on is designed to build or destroy patronage empires.

This makes Maharashtra sound more like an idealized Western political system than perhaps it should. It is not, after all, pluralist in quite the sense that the US claims to be (though it may have more resemblance to the US system as it in fact is -- somewhat pluralist but closer to the elite-domination end of the spectrum). Furthermore, the EGS is not a panacea for rural poverty. It may have put a floor under rural incomes (in contrast to BB, where real income for agricultural labor has fallen in recent decades as this poor of residual rural labor has grown) and may even have raised rural incomes slightly, but it has scarcely solved the problems of rural poverty. But the social/economic/political infrastructure of Maharashtra has provided a kind of security for the rural poor that is simply nowhere in sight in BB, and it has also provided an integrated dynamic of rural development that is just not present in BB.

IV. Implications for Rural Development Policy

In Maharashtra, or certainly WM, the themes of rural development appear to be capitalist agriculture, cooperative-based politics, interlinked institutional infrastructure to support development and (to a more modest extent) social responsibility. In BB, the predominant themes are a semi-feudal agriculture, a predatory politics reinforced by a functionally diffuse institutional infrastructure and a lack of social responsibility. The key difference in terms of rural development strategy would seem to the emergence of an infrastructure of (more or less) functionally discrete institutions in WM that do not get sucked into and perverted by the maelstrom of patronage politics but instead can actually deliver goods and services to promote rural development. What then might be suggested in the way of rural development policy to improve things in BB?

In Bihar there are some indications of an agricultural transformation and the mobilization of a constituency to support it (with Karpoori Thakur and the Lok Dal being the premier case in point), but the process is at best slow and seems to have stalled indefinitely after the brief Lok Dal surge to power in the late 1970s. Can the momentum be regained, whether with the Lok Dal or with a reoriented Congress Party that seeks to mobilize "progressive farmer" strata? There are also signs of political consciousness and organization among the landless, particularly Harijan groups. Can they claim a niche in the political system? At present, it would appear that any solutions to the dilemmas of rural stagnation would have to stem out of political action from below -- from the middle and lower orders breaking into the political arena and forging a place for themselves, from which they could demand policy and implementational changes that would begin to seriously deliver the goods for rural development. Or to put it the other way, there seems virtually no chance of a leadership emerging at the top with a serious program for developing the state. If anything, the factional intercaste and intracaste struggle among the "forwards" that so debilitated the state from the 1950s through the early 1970s is today even more intense. This is not the sort of environment to nurture rural development.

In Bangladesh, on the other hand, there would seem to be more room for maneuver. Unlike Bihar, where rural elites with large landholdings tie directly into the upper echelons of state power, in Bangladesh the dominant elements in the countryside are smaller in scale -- in Bihar they would be middle peasants, a situation partly due to land pressure fragmenting holdings and partly to the departure of so many Hindu zamindars after the partition, thereby leaving the field to the smaller fry. Thus while the state does need local elites to help maintain order, it is not beholden to them in quite the same way as in Bihar, for they themselves do not loom so large as a class on the wider scene of state politics.

One of the consequences of this relative separation is that the state has somewhat more autonomy to pursue reformist rural development policies. Nothing on the order of land reform is in the cards, to be sure, but the state does have more latitude to tolerate and even modestly support non-governmental organizations (NGOs) that in Bihar would be denounced as unacceptably radical. Hence groups like the Bangladesh Rural Advancement Committee (BRAC) and Proshika can advance "conscientization" agendas in their rural development work, and have been doing so in some cases for fifteen years and more. And quasi-parastatals like the Grameen Bank (that now has several hundred thousand borrowers) can lend to the rural poor without being taken over by the rural rich. Altogether, it is estimated that some 500,000 rural households -- about seven percent of the total -- are covered by one or another of the two hundred or so NGOs now operating in Bangladesh. By no means all these NGOs are embarked on agendas like BRAC's or Proshika's, but a good many of them are, and so far the regime has found them acceptable.

It is also possible to encourage food for work enterprises like those of CARE and the World Food Programme that are administered with a creditable degree of probity (though the variant on food for work that is run by the government itself is quite different in this respect). In recent years there is even some inferential evidence that rural agricultural wages have gone up,(17) a development which would have to be due in part to these food for work activities and one which would never be tolerated in rural Bihar.

The state has also been promoting a decentralization scheme in the form of its upazila parishad system,(18) in which popularly elected councils for the first time are gaining control of centrally funded development programs in their regions. This effort at present is caught in a sort of pincers, with one tong representing an effort by the regime to build up a rural power base by decentralization the other an effort on the part of the bureaucratic cadres to retain their centralized power. Even to the extent that the upazila initiative succeeds, it can be seen as an attempt to cement ties between the state and rural elites, which I would suspect is in large part its intended design. But the upazila parishads can also offer a venue for other groups to enter the political arena and become involved, even to take over with their superior numbers and voting power. And given the relatively free reign enjoyed by NGOs in rural Bangladesh, it is more than remotely conceivable that reformist NGOs could mobilize large segments of the rural poor to mount such a takeover at local level.

Moving into terrain as yet uncharted, it might be possible to organize jute processing and marketing cooperatives along lines somewhat similar to those pursued by the sugar cooperatives in Maharashtra (or perhaps similar to the Anand milk scheme in Gujarat). Jute is a cash crop that is not consumable by the growers and which needs considerable processing before it can be sold. Much of the profit in the crop is made by the farias or middlemen who buy from the farmers and sell to processors at higher prices. The farias could be cut out, cooperative mills could be set up to process the jute into burlap and carpet backing (the major end products today), and a federation of the processing cooperatives could market it in the international arena. There would be some serious difficulties here, if only because jute has lately faced a declining demand on world markets, but the situation is far from hopeless.

In short, I believe that there are at least the possibilities there in rural Bangladesh to build the kind of rural institutional infrastructure and linkages that exist in Maharashtra. Much will depend on how the state assesses its opportunities, how various elites in the system act and react (not least the army, certainly), and what policies the international donor community pursues. At best, even a partial replication of the Maharashtra experience will take a long time, probably longer than the forty years and more it took in its original setting. But the potential is there, if not for a restoration of the "Golden Bengal" of mythical antiquity, at least for a significant improvement over the current scene.

Table 1
Total Population and Largest Cities, 1981/1991
Urban/Rural breakdownMaharashtra Bangladesh Bihar
'81'91 '81'91 '81'91
Total population62.8m78.7 87.1m106.3 69.9m86.3
Urban population22.0m30.5 13.2m20.9 08.7m11.4
Rural population40.8m48.2 73.9m85.4 61.2m74.9
Percent urban35.0%38.7 15.2%19.6% 12.4%13.2%
Percent rural65.0%61.3% 84.8%80.4% 87.6%86.8%

City rank and sizeMaharashtra Bangladesh Bihar
'81'91 '81'91 '81'91
Largest city8.23m12.57m 3.44m6.95m 0.92m1.10m
Second largest1.69m2.49m 1.39m2.35m 0.68m0.84m
Third Largest1.55m1.66m 0.62m1.00m 0.67m0.82m
Fourth largest0.65m0.72m 0.17m0.54m 0.50m0.61m
Fifth largest0.52m0.62m .16m0.21m 0.26m0.42m
Sixth largest0.43m0.59m 0.16m0.21m 0.25m0.29m
Seventh largest0.39m0.43m 0.16m0.19m 0.22m0.26m
Eighth largest0.35m0.42m 0.15m0.16m 0.19m0.24m
Ninth largest0.32m0.39m 0.13m0.15m 0.18m0.22m
Tenth largest0.27m0.36m 0.13m0.14m 0.15m0.20m

Table 2
Literacy in Maharashtra and Bengal/Bihar
as a percentage of population, 1931-1991

(Boundaries as of 1981)

Male literacy
YearMaharashtraWest Bengal BangladeshBihar

Female literacy
YearMaharashtraWest Bengal BangladeshBihar
195191111 4
a For Bangladesh, the census year was 1974 rather than 1981.

Table 3
Principal Ethnic Groups in Bihar, Maharashtra and Bangladesh

(groups with >2% of population int he early 1970s 1)

Regions - Bihar Maharashtra Bangladesh
All Hindu Castes 84 88 14

Brahman - 5
Bhumihar - 3
Rajput - 4
Brahman - 4
Maratha -32


Yadav - 11
Kurmi - 4
Koiri - 4
Teli - 3
Mali - 3


All castes - 14
Chamar - 4
Dusadh - 4
Musahar - 2
All castes 2 - 13
Mahar 2 - 9

All Muslims 14 8 85
Other groups 3 4
Grand total 100 100 100

1. Sources used: Mainly 1931 census for Hindu castes, with boundaries adjusted to conform with 1981 areas; 1971 census for Muslims (1974 for Bangladesh).
2. Includes both Hindus and Buddhists.

Political-Economic Development Linkages in Bihar & Bangladesh, Mid-1980s

Political-Economic Development Linkages in Rural Western Maharashtra: Ahmednagar District Mid-1980s

Principal Ethnic Groups in Bihar, Maharashtra, and Bangladesh


1. Needless to say, my indebtedness is enormous, to both institutions and individuals in the United States and South Asia. Full acknowledgments will have to come later on when the work is more complete, but for now I would like to thank the American Institute of Indian Studies, the Fulbright research program, the Social Science Research Council, the Rural Development Committee at Cornell University, and Bucknell University for their support at various stages along the way.

2. I take refuge in this term, which is I believe reasonably accurate, but at the same time sufficiently vague that it is possible to avoid the "mode of production" debate.

3. The term "state" presents a mild conceptual confusion here, in that for Bihar the government is a state which is part of a larger central government, whereas in Bangladesh the state is of course a national government in its own right. In terms of the rural political economy, however, the state has virtually the same relationship to the countryside for both cases. The maintenance of state power today means continuation of a Congress ministry in Patna and of the Ershad regime in Dhaka.

4. It is more difficult to deal with regions in Maharashtra than for Bihar and Bangladesh, in large part because while the latter have been more or less identifiable since the division of the Bengal presidency in 1912 (and in many ways date back long before that), Maharashtra is a much more recent creation, having been formed in its present shape only in 1960. The focus of this paper is on the western Deccan lava of Maharashtra that is the major sugar producing area of the state; this consists of the Pune division's six districts plus three from Bombay division (Nashik, Dhule and Jalgaon), or altogether nine of the state's present 27 districts, which are collectively often called Western Maharashtra. All of the nine save Kolhapur and a portion of Sangli districts (which were former princely states) were part of the Bombay presidency during the British period, thus giving them a common history not shared by the Marathwada and Vidharbha regions to their east. For Bihar I should note that the present analysis applies basically to the Gangetic plain region of the state, not to the largely tribal area of Chotanagpur.

5. Even gur-making is too complex to be done at home. The process requires an investment of several thousand dollars and perhaps a dozen or more employees working at any one time.

6. Some perspective here might be gained from reflecting on the awesomely high subsidies given to water supplied in the western United States by the U.S. Bureau of Reclamation.

7. Much of the irrigation in WM is through gravity flow channels that do not require pumps, but a good part comes from wells, which are charged primarily through underground percolation from unlined channels which leak a good deal of their water into the adjacent soil. For farmers benefiting from this indirect water supply, electricity for pumps is a major consideration.

8. In the summer of 1987, for example, the world price of sugar was about US 6¢/lb., while inside India it was the equivalent of around US 25¢. At times, of course, the world price might be in excess of the internal Indian price, but in recent years the international price has tended to stay quite low.

9. There are other programs patterned on EGS that have been initiated in recent years by the center, in particular the Rural Landless Employment Guarantee Scheme and the National Rural Employment Programme, but neither thus far is more than a pale imitation of the original.

10. Buying these would-be "progressive farmers" off for the Congress would presumably mean providing them with some of the inputs and technologies that their counterparts elsewhere in India have been getting for years and even decades from the government, so such "bribery" is another way of saying instrumental development-oriented politics. And indeed such a politics might lead to a significant degree of development in Bihar, were it not for the huge dead weight of the upper gentry castes controlling the overall direction of political life there.

11. Altogether the various food for work programs in Bangladesh are estimated to generate upwards of 70 million work days annually.

12. That effect has been uneven, with those better off getting a larger share, but then what place outside of China in the Cultural Revolution has not? The point is that those in the lower income deciles have been getting something tangible from the rural development process, whereas in BB this has evidently not been the case.

13. Reconstructing the present area of Maharashtra from compilations of the British period is a tedious task, for it involves adding in and subtracting out a number of districts from different British provinces as well as a large group of princely states. Bihar and Bangladesh are a great deal easier to deal with in this respect.

14. To some extent these observations may be more revealing of north-south (or Aryan-Dravidian) differences in India than east-west divisions.

15. Ironically, Huq himself devoted much of his energy to keeping Hindus and Muslims united, but that is not what he is remembered for.

16. Interestingly, rural violence, though far from absent in Bangladesh, appears much attenuated over what it was in the early 1970s, when the well-armed remnants of various freedom fighter groups roamed the countryside.

17. Recent World Bank data show a 2.4% annual increase in real rural wages over the 1979-1986 period, a rather startling development after the steady downward trend during the 1970s. Studies by other donors indicate similar trends. Personally, I must confess some skepticism here, for the news seems too good to make sense, in view of what is known about the relative increases in the labor force and available employment.

18. Upazilas ("next to district" in Bengali) are the old thanas (analogous to tehsils, talukas, anchals or blocks in those areas of India outside the old Bengal Presidency).

Back to the top.

Return to Peasant Symposium program