Jump to ContentJump to Main Navigation
Hunger and Public Action$

Jean Drèze and Amartya Sen

Print publication date: 1991

Print ISBN-13: 9780198283652

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0198283652.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 20 October 2021

China and India

China and India

(p.204) 11 China and India
Hunger and Public Action

Jean Drèze (Contributor Webpage)

Amartya Sen (Contributor Webpage)

Oxford University Press

Abstract and Keywords

Taking advantage of the similarities between China and India when they began their development planning, an assessment is made of their achievements and how eventually China had greater success than India in fighting endemic deprivation. The Chinese experience is seen as a success story of support‐led security solving the problem of chronic hunger. This is contrasted with the unsuccessful experience with the Chinese famine. The last part focuses on Kerala, which, although one of the poorer states in India, has managed to achieve high levels of life expectancy.

Keywords:   China, Chinese famine, India, Kerala, persistent undernutrition, public support, support‐led security

11.1 Is China Ahead?

When development planning began in China after the revolution (1949) and in India after its independence (1947), both countries were starting from a very low base of economic and social achievement. The gross national product per head in each country was among the lowest in the world, hunger was widespread, the level of illiteracy remarkably high, and life expectancy at birth not far from 40 years. There were many differences between them, but the similarities were quite striking. Since then things have happened in both countries, but the two have moved along quite different routes. A comparison between the achievements of China and India is not easy, but certain contrasts do stand out sharply.

Perhaps the most striking is the contrast in matters of life and death. Life expectancy at birth in China appears to be firmly in the middle to upper 60s (close to 70 years according to some estimates),1 while that in India seems to be around the middle to upper 50s.2 The under‐5 mortality rate, according to UNICEF statistics, is 47 per thousand in China, and more than three times as much in India, viz. 154.3 The percentage of infants with low birth weight in 1982–3 is reported to be about 6 in China, and five times as much in India.4 Analyses of anthropometric data and morbidity patterns confirm that China has achieved a remarkable transition in health and nutrition.5 No comparable transformation has occurred in India.

Things have diverged radically in the two countries also in the field of elementary education. The percentage of adult literacy is about 43 in India, and around 69 in China.6 If China and India looked similar in these (p.205) matters at the middle of this century, they certainly do not do so now.7

China and India

Fig. 11.1 Life expectancy at birth in China and India

The comparison is not, however, entirely one‐sided. There are skeletons in China's cupboard—millions of them from the disastrous famine of 1958–61. India, in contrast, has not had any large‐scale famine since independence in 1947. We shall take up the question of comparative famine experience later on (in section 11.3), and also a few other problems in China's success story (sections 11.4 and 11.5), but there is little doubt that as far as morbidity, mortality and longevity are concerned, China has a large and decisive lead over India.8

(p.206) What has brought about that lead is a matter of very considerable interest, and to that issue we now turn.

11.2 What Put China Ahead?

In the last chapter we distinguished between ‘growth‐mediated security’ and ‘support‐led security’. Is China's success based mainly on rapid economic growth, or is it primarily the result of developing a vast system of public support? It might not be immediately obvious what has been the main factor behind China's lead over India. Of course, China has had a much more extensive system of public delivery of food, health care, and education. But, according to widely quoted standard statistics, it is also credited with having had a much faster rate of economic growth than India. We have to scrutinize both sets of information.

Comparison of China's and India's rates of growth of per‐capita GNP has attracted a good deal of attention among experts, and the difficulties of comparison are indeed formidable. The consensus has been in the direction of accepting a much higher growth rate in China than in India, though the difference appears to be somewhat smaller after corrections are made to account for the different bases of the respective estimates. There are also some non‐comparabilities arising from differences in the way in which the two economies function (e.g. in the role of the service sector, which is much larger in India and has been growing fast), and the authors of these studies have warned us to be careful in interpreting the comparative estimates.9

World Development Report 1988 gives the growth rate of GNP per head during 1965–86 as 1.8 per cent in India and 5.1 per cent—nearly three times as large—in China. If the comparative growth picture is really something like that, it would be natural to assume that the Chinese achievement in matters of life and death must have been substantially ‘growth‐mediated’. Growing at 5.1 per cent per year over decades would, for one thing, make a country's per‐capita income a great deal higher than what would result from hastening slowly at 1.8 per cent, and that—as the arguments of the last chapter indicated—could make a lot of resources potentially available for public support in addition to enhancing average personal affluence.

These comparative growth figures do not, however, bear any scrutiny. Indeed, the same World Development Report 1988 gives the Chinese per‐capita GNP in 1986 as $300 and that of India as $290. If the figures of GNP level and GNP growth are put together, it would appear that in 1965 India had about twice the per‐capita GNP of China (about $200 for India and $106 for China at 1986 prices). That is, in order for the Chinese GNP per head to have grown at (p.207) 5.1 per cent while India's grew at 1.8 per cent, and for the two to have ended up with their respective 1986 GNP figures, the Chinese per‐capita GNP in the middle 1960s would have had to be only a little over a half that of India.10

That inference would be contrary to all available estimates of Chinese and Indian GNP for the 1960s. It would have also made China then a great deal poorer than the poorest country in the contemporary world.11 In fact, Simon Kuznets's estimates indicated a similar GNP per head in 1958 for China and India, with ‘product per head’ about 20 per cent higher in China.12 The story of the Chinese GNP per head growing about three times as fast as the Indian simply does not bear empirical scrutiny.

In fact, the World Bank's own World Tables 1987 puts the GNP per head in the mid‐1960s, specifically 1966, at $110 for China and $90 for India, i.e. 22 per cent higher in China. World Tables 1987 also gives the 1986 GNP per head of China and India as $300 and $270 respectively. Putting their 1966 and 1986 figures together indicates implicit growth rates that are very similar for the two countries (5.1 per cent per year for China and 5.6 per cent in India). That is a very different picture from that of China's growth rate of 5.1 per cent per year being nearly three times as high as India's 1.8 per cent, as given by the more widely used publication of the same World Bank—the World Development Report. The picture is confused and confounded, to say the least.

One should not make heavy weather of these inconsistencies. We do know that international comparisons of GNP are plagued by the usual complications of exchange‐rate variations and changes in relative prices.13 But the complete lack of relation between GNP levels and growth rates in World Development Report's account of China's and India's respective performance makes it hard to put much reliance on the respective growth statistics.14 In particular, the (p.208) story of a growth rate in China very much higher than that of India (5.1 per cent against India's 1.8 per cent) is difficult to firm up. It is hard to see the Chinese attainment in health and nutrition as primarily a ‘growth‐mediated’ success.

In fact, it seems fairly clear that the Chinese growth rate was not radically higher than that of India before the economic reforms of 1979, by which time the tremendous surge ahead in health and longevity had already taken place.15 In the pre‐reform period, agricultural expansion in particular was sluggish in China, as it was in India, and the dramatic reduction in hunger and undernourishment and expansion of life expectancy in China were not ushered in by any spectacular rise in rural incomes or of food availability per head. As Judith Banister notes: ‘It also appears that the quantity of food produced per capita and the quality of the Chinese diet did not improve between 1957 and the late 1970’s. . . . annual per capita grain production through 1977 was about the same as in the late 1950's: it averaged 301 kilograms in 1955–57 and 305 kilograms in 1975–77.'16

This is indeed the crucial point. The Chinese level of average opulence judged in terms of GNP per head, or total consumption per capita, or food consumption per person, did not radically increase during the period in which China managed to take a gigantic step forward in matters of life and death, moving from a life expectancy at birth in the low 40s (like the poorest countries today) to one in the high 60s (getting within hitting distance of Europe and North America).

Since the far‐reaching economic and social reforms introduced in 1979 (with much greater use of market‐based incentives), the Chinese growth rate has been fast—very much faster than earlier. But this has not been a period of further reduction in mortality. In fact, quite the contrary. The death rate in China reached, according to official statistics, its lowest level in 1978, just before the reforms, and went up in the period following the reforms, precisely when the growth of output and income accelerated impressively. On the lessons of the post‐reform period in China, we shall have to say more (see section 11.4).

At the moment we only note the fact that the great increase in longevity and reduction in mortality took place in China before the reforms, during a period of fairly moderate economic growth, whereas the post‐reform period of rapid growth has witnessed no further rise and possibly some deterioration in survival chances. China's remarkable achievements in matters of life and death (p.209) cannot in any way be ascribed simply to a strategy of ‘growth‐mediated’ security.

As far as support‐led security is concerned, the Chinese efforts have been quite spectacular. The network of health services introduced in post‐revolutionary China in a radical departure from the past—involving cooperative medical systems, commune clinics, barefoot doctors, and wide‐spread public health measures—has been remarkably extensive.17 The contrast with India in this respect is striking enough. It is not only that China has more than twice as many doctors and nearly three times as many nurses per unit of population as India has.18 But also these and other medical resources are distributed more evenly across the country (even between urban and rural areas), with greater popular access to them than India has been able to organize.19

Similar contrasts hold in the distribution of food through public channels and rationing systems, which have had an extensive coverage in China (except in periods of economic and political chaos, as during the famine of 1958–61, on which more presently). In India public distribution of food to the people, when it exists, is confined to the urban sector (except in a few areas such as the state of Kerala where the rural population also benefits from it, on which, too, more presently). Food distribution is, in fact, a part of a far‐reaching programme of social security that distinguishes China from India.20 The impact of these programmes on protecting and promoting entitlements to food and basic necessities, including medical care, is reflected in the relatively low mortality and morbidity rates in China.

The contrast between China and India in public distribution systems and in social security programmes is certainly very striking, and it is plausible to see China's success story as one of support‐led security. The growth‐mediated interpretation would be much harder to defend, especially since there are so many uncertainties about what GNP growth rates China did, in fact, achieve. We shall come back to the growth question again in section 11.4 when we review the impact of the recent experience of high economic growth in China (p.210) since the reforms. But before that we examine the big blot in China's past record, viz. the famine of 1958–61.

11.3 The Chinese Famine and the Indian Contrast

The Chinese famine of 1958–61 followed the débâcle of the so‐called Great Leap Forward that was tried out from late 1957 onwards.21 While the failure of the Great Leap Forward came to be widely recognized after the initial euphoria, the existence of the famine oddly escaped open scrutiny and even public recognition, until very recently.22 This is particularly interesting given the monumental scale of the famine—arguably the largest in terms of total excess mortality in recorded history.

Comparing actual mortality with pre‐famine mortality yield remarkably high figures of extra deaths in this famine. Estimates of extra mortality vary from 16.5 million to 29.5 million.23 These figures are extraordinarily large. For example, the excess mortality in the last Indian famine, viz. the so‐called Great Bengal famine of 1943 (occurring four years before independence), is estimated to be about 3 million.24 In the scale of ‘extra deaths’ the Chinese famine was, thus, about five to ten times as large as the largest famine in India in this century.25

Many things are still uncertain about the causation of the famine, but it is clear that there was an enormous collapse of agricultural output and income. Food availability decline certainly played an important part in the genesis of hunger that gripped China for three years. Taking the average national grain output per capita of 308 kg. in 1956 and 1957 as the point of reference, the 1959 output was 17 per cent below this and by 1960 the per‐capita grain output was as much as 30 per cent down. Even in 1961, the shortfall vis‐à‐vis the 1956–7 (p.211) average was 28 per cent, and the pre‐famine figure was not reattained until the latter half of the 1960s.26

Further, some regions suffered particularly serious declines and the sharp differences between food availability in the different regions continued through the period of distress.27 For example, in 1960, while the provinces of Heilongjiang and Yunnan had respectively 229 and 209 kg. of grains per head for their rural population, the corresponding figures for Henan, Sichuan, and Hebei were respectively 143, 137, and 122 kg. The contrast between the rural and urban areas was also striking (for example, between 288 kg. in urban Hebei as against 122 in the rural areas of that province).28

Public distribution at the local level was comprehensively disrupted. The problem for the rural areas was made much worse by the sharp increase in state procurement of foodgrains, and the rural communes were in many cases desperately short of food. In addition to that, there were also remarkable inequalities in the distribution of whatever food was available. This applied not only between the regions and between urban and rural areas, but even within a given rural region. Some provinces evidently suffered from much sharper intraregional inequalities than others did, with correspondingly higher mortality. For instance, rural Sichuan and Henan suffered from much greater death rates than rural Hebei which did not have, on the average, any more food. Particularly, there was a great deal of wastage and excess consumption in particular ‘commune mess halls’.29

The Chinese famine of 1958–61 was closely linked with policy failures—first in the débâcle of the Great Leap Forward, then in the delay in rectifying the harm done, and along with that in accentuating distributional inequalities through enhanced procurement and uneven sharing. The remarkable aspect of the famine is its continuation over a number of years without an adequate recognition of the nature of the crisis (and without leading to the necessary changes in public policy).

This is one respect in which India's record since independence must be seen to be very much superior. The fact that there has been no large‐scale famine in India since independence is a positive contrast with the Chinese experience. The contrast is particularly interesting when account is taken of the fact that there have been several alarming dips in food output and availability in India over the same period (the latest being in the drought of 1987), and that on many occasions the entitlements of large parts of the population have been severely threatened both directly and indirectly (particularly through employment declines associated with droughts or floods).

(p.212) The Indian system of famine prevention was discussed in Chapter 8 of this book. There are, as was argued, two different features involved in the system. One is a worked‐out procedure for entitlement protection through employment creation (usually paying the wages in cash), supplemented by direct transfers to the ‘unemployable’. The origins of this procedure go back to the 1880s and the Famine Codes of the late nineteenth century, through a number of important developments (including the use of the public distribution system to stabilize food prices) have taken place since independence. The other part is a political ‘triggering mechanism’ which brings the protection system into play and indeed which keeps the public support system in a state of preparedness. It was this triggering mechanism that was lacking in the famine prevention system of British India after the Famine Codes were set up. In the Bengal famine of 1943, not only were the Famine Codes not invoked, that was indeed a deliberate decision of the government.30

On the other hand, given the political system of post‐independence India, it is extremely hard for any government in office—whether at the state level or at the centre—to get away with neglecting prompt and extensive anti‐famine measures at the first signs of a famine. And these signs are themselves more easily transmitted given India's relatively free media and newspapers, and the active and investigative role that journalists as well as opposition politicians can and do play in this field. The adversarial participation of newspapers and opposition leaders is, as we have discussed earlier, an important part of the Indian famine prevention system. It yields a rapid triggering mechanism and encourages preparedness for entitlement protection.

The contrast with China is striking primarily in the second respect. Given its system of public distribution, China did not lack a delivery and redistribution mechanism to deal with food shortages as the famine threatened in 1958 and later. Despite the size of the decline of food output and the loss of entitlement of large sections of the population, China could have done a much better job of protecting the vulnerable by sharing the shortage in a bearable way.

What was lacking when the famine threatened China was a political system of adversarial journalism and opposition.31 The Chinese famine raged on for three years without it being even admitted in public that such a thing was occurring, and without there being an adequate policy response to the threat. Not only was the world ignorant of the terrible state of affairs in China, even the population itself did not know about the extent of the national calamity and the extensive nature of the problems being faced in different parts of the country.

Indeed, the lack of adversarial journalism and politics hit even the government, (p.213) reinforcing the ignorance of local conditions because of politically motivated exaggeration of the crop size during the Great Leap Forward and the fear of local leaders about communicating their own problems. The pretence that everything was going all right in Chinese agriculture and rural economy to a great extent fooled the national leaders themselves. ‘Leaders believed in 1959–60’, as Bernstein puts it, ‘that they had 100 MMT more grain than they actually did.’32 This misconception was crucial in keeping down Chinese imports of foodgrains, which fell to virtually nothing in 1959 (about two thousand tons compared with 223 thousand tons in 1958), and stayed incredibly low in 1961 (66 thousand tons), before jumping to 5.8 million tons in 1961 as the fact of the famine and the agricultural débâcle became at long last clear. Chinese exports of foodgrains, similarly, peaked in 1959, and stayed high in 1960, before beginning to come down in 1961. The Chinese net exports of cereals rose from 1.9 million tons in 1957, to 2.7 in 1958, 4.2 in 1959, and 2.7 in 1960—as the famine devastated the lives of tens of millions of people across the country.33

The misinformation and misreading also led to a sharp increase in the extent of food procurement from the rural areas. The percentage net procurement out of total output went up from 15 to 17 per cent in 1956 and 1957 to 21 per cent in 1958 and 28 per cent in 1959.34 The rural Chinese—hit by a production decline—were hit again by having to part with a larger proportion of the reduced output as procurement by the state.

The misinformation also contributed to the non‐revision of production and distribution policies and to the absence of any emergency entitlement‐protection programme.35 Aside from the government's informational inadequacy, which made its own assessment of the situation disastrously wrong, the absence of an adversarial system of politics and journalism also meant that there was little political pressure on the government from any opposition group and from informed public opinion to take adequate anti‐famine measures rapidly.

We end this section with three interpretative remarks. First, as was discussed (p.214) in Chapter 8, famine prevention is an important achievement of India, and there is something to learn from that experience in this famine‐ridden world. The fact that even post‐revolutionary China, with its outstanding record of entitlement promotion and enhancement of living conditions, could fall prey to a gigantic famine indicates that the lesson may be far from negligible. In fact, the precise feature of absence of adversarial politics and open journalism that may have contributed to the occurrence, magnitude and duration of the Chinese famines of 1958–61 are also present in most sub‐Saharan African countries today. While the political systems are quite different, this feature of absence of political opposition and free journalism in African politics is a cause of famine vulnerability in Africa as it was in China at the time it had its own disaster. Also, greater tolerance of criticism and more open journalism in China would have a positive effect on helping to make China secure against the kind of political and economic crisis that ushered in the famines of 1958–61. But unfortunately political democratization in China has not really kept pace with the speed of economic liberalization (on which more presently).

Second, as India's experience shows, open journalism and adversarial politics provide much less protection against endemic undernutrition than they do against a dramatic famine. Starvation deaths and extreme deprivation are newsworthy in a way the quite persistence of regular hunger and non‐extreme deprivation are not. Endemic hunger may increase the morbidity rate and add to the mortality rate (in these respects India's performance continues to be quite awful), but that is primarily a statistical picture rather than being immediately palpable and—no less importantly—being ‘big news’. To bring endemic deprivation into the fold of news reporting and to make it a major focus of political confrontation are inherently more difficult tasks, and seem to have been largely beyond the normal activities of journalists and politicians in India.36 That situation could change (there are some signs of that already), and this is clearly a field in which there is scope for the public to play a very creative role in India. But as things stand, the Chinese political commitment—not unrelated to the ideological predispositions of the Chinese political system—seems to have served the country well for combating endemic deprivation, despite its failure as a defence against famines.

Finally, it is important to note that despite the gigantic size of excess mortality in the Chinese famine, the extra mortality in India from regular deprivation in normal times vastly overshadows the former. Comparing India's death rate of 12 per thousand with China's of 7 per thousand, and applying that difference to the Indian population of 781 million in 1986, we get an estimate of excess normal mortality in India of 3.9 million per year. This implies that every eight years or so more people die in India because of its higher regular death rate than died in China in the gigantic famine of (p.215) 1958–61.37 India seems to manage to fill its cupboard with more skeletons every eight years than China put there in its years of shame.

Table 11.1 China Since the 1979 Reforms

Gross value of output (index)

Death rate (index)

Female‐male ratio (value)





















































Source: People's Republic of China, Statistical Yearbook of China 1986 (in English); 1987 (in Chinese).

11.4 Chinese Economic Reforms: Opulence and Support

The economic reforms introduced in China in 1979 have now gone through nearly a decade of practice, and it is possible to begin assessing some of their impacts. On the side of commodity production, there is little doubt that the Chinese economy has surged ahead in response to market incentives, and the agricultural sector has really had—at long last—a proper ‘leap forward’. As Table 11.1 indicates, the gross value of agricultural output doubled between 1979 and 1986 (a growth rate of more than 10 per cent per year). Growth in the agricultural sector has also kept pace with industrial expansion, which is in fact quite remarkable in itself, particularly since industrial expansion has been very fast as well. There may be some questions about the exact figures (there might have been incentives to understate output prior to the reforms), and it is not easy to be certain of the exact growth rates achieved. There has also been a slow‐down after the initial leap, as well as some worry that the production of food crops has not kept up with the expansion of other types of production. Nevertheless, taking everything into account, there can be little doubt that the economic reforms have been quite remarkable in expanding the supply of nutrients in China as well as agricultural outputs and incomes in general.38

The economic reforms can be and have been questioned from many points of view. The fact that the reforms have led to an inflationary situation with price (p.216) rises destabilizing the consumers' equilibrium has been widely acknowledged, and this has caused some rethinking on how far and how fast to go on the path of economic change. There have also been fears that the price mechanism, while successful in raising total outputs and incomes, may increase inequalities in the distribution of incomes. But, again, even after taking note of these qualifications, it is quite clear that the average opulence of the Chinese population, especially the rural population, has expanded greatly since the reforms.

Even in the reduction of poverty, calculated in terms of personal incomes, a great deal has been recently achieved. It appears that the number of rural Chinese below the poverty line of 200 yuan in 1986 prices has fallen from 200 million in 1979 to 70 million in 1986.39 That is a striking decline of which there are few parallels. If ‘growth‐mediated security’ were the chief means of promotion of longevity in China, this post‐reform experience should have provided an excellent basis for further enhancement of life expectancy at a rapid rate.

In fact, however, this has not occurred, as we noted earlier. The death rate in China, rather than declining rapidly, seems to have gone up after the reforms, as Table 11.1 indicates. Indeed, in no year since 1979 has the death rate—as given in Chinese official statistics—been lower than that achieved by 1978 and 1979, viz. 6.2 per thousand. The death rate, as reported in the Statistical Yearbook of China, went up to 7.1 by 1983, and even after coming down, it has hovered around 6.6 and 6.7 in 1985 and 1986.

There is much room for doubt about the correctness of the official Chinese mortality data. It must be particularly noted that an increase in the coverage of mortality statistics may have the effect of raising the reported death rate. It is quite possible that at least some part of the apparent increase in mortality rates after 1979 is connected with better coverage of death data. Also, we have to take note of the changing age composition of the Chinese population when interpreting overall death rates.40 But even after note has been taken of these factors, there is evidence of an increase in forces of mortality since the reforms compared with what had been achieved before them.41 The downward trend in mortality which made China reach truly unusual levels of longevity (given its low per‐capita income) has been at least halted, and possibly reversed.

The effect of the changing age composition can be eliminated by looking not at the crude death rate, which is in effect a simple average, but at the life expectancy figures. Life expectancy is estimated by using ‘endogenous weights’ (in the sense that the population in different age groups is estimated (p.217) by using the mortality rates in the previous age groups), and does not depend on the actual age structure of the existing population. Thus, concentrating on life expectancy gives us a good idea of what we are looking for.

Table 11.2 Life and Death in China, 1978–1984


Crude death rate (per thousand): Banister's adjusted estimates

Life expectancy at birth (years): Banister's estimates

Infant mortality rate (per thousand)

Banister's adjusted estimates

Yang and Dowdle indirect estimates


































Source: Banister (1987) Table 4.18 and Yang and Dowdle (1985).

Judith Banister's estimates of life expectancy at birth are given in Table 11.2. These suggest a steady decline in life expectancy since 1978 (up to and including 1984, which is the last year in Banister's series). The fall is moderate, though firm and consistent, but it has to be judged particularly as a contrast with steadily declining mortality rates and expansion of life expectancy up to the late 1970s (with the exception, of course, of the period of the famine of 1958–61, discussed earlier). The real issue is the slowing down of social progress just when overall economic growth has quickened.

While the reduction of life expectancy is fairly moderate, the rise of infant mortality according to some estimates appears to be sharp (see Table 11.2). This is so according to both Banister's estimates and also the indirect estimates made by Yang and Dowdle (1985) on the basis of a fertility survey questionnaire. The extent to which this is connected with China's enforced population policy introduced also in 1979—especially the insistence on a one‐child limit in many parts of the country and a two‐child limit elsewhere—is not crystal clear, but there is strong circumstantial evidence in that direction.42 One of the sinister signs is a decline in the reported birth ratio of females to males, and this can reflect infanticide, or at least death due to severe neglect, of female children, with their births as well as deaths remaining unregistered.43 The recent (p.218) relaxation of the one‐child limit when the first child is a girl may have been the result of recognizing the prevalence of this problem.

Table 11.3 Gender Differential in Mortality in China, 1978–1984


Life expectancy at birth (years)

Infant mortality per thousand



Female advantage



Female advantage


















































Source: Banister (1987), Table 4.12.

No matter what role the population policy may have played in this, there is considerable evidence that the mortality picture has possibly darkened for girls in comparison with boys in recent years. In fact, Banister's estimations suggest that in the post‐reform period while male infant mortality has continued to fall (though only quite slowly), it is female infant mortality that has apparently had an upward jump (see Table 11.3). Similarly and correspondingly, male life expectancy at birth has continued to rise (again, very moderately since the reforms), but female life expectancy reached a peak in the pre‐reform year of 1978 and has fallen apparently since then. These estimates are speculative and must not be taken too seriously, but altogether there is much evidence of (1) a slow down or a halt in the steadily improving survival chances of both men and women, and specifically children, and (2) on top of that, an evident increase in gender bias, specifically affecting female children.

The increase in gender bias, if confirmed, would no doubt relate (at least partly) to recent population policies. However, enforced limitation of population size, since it is in principle ‘gender neutral’, can have a devastating effect on female children only when there is already a strong parental preference for male children, and ultimately we have to be concerned also with the causal roots of that preference.

Such male‐preference is not, of course, a new thing in China, but it is arguable that the responsibility system may itself have contributed a little to the undermining of the position of women. As was argued in Chapter 4, there is considerable evidence that the involvement of women in so‐called ‘gainful employment’ tends to reduce gender bias against females. In this respect, the communal form of agriculture used in pre‐reform China provided much easier scope for female ‘gainful’ involvement, and the proportion of women in such employment had risen quite radically in the 1960s and 1970s. However, with the new responsibility system, Chinese agricultural production has become (p.219) more family‐based, with the usual division of labour that tends to place women in activities of the typical ‘household’ kind.44

This can indeed be an influence towards worsening the position of women in ‘cooperative conflicts’, and through a general regression of women's economic position and social status, can also strengthen the anti‐female bias in the caring of children. It is not obvious that this type of effect, which—if important—is most likely to be so only in the long run, could have had any role already in the Chinese rural society. But no matter what view is taken of that question (and more generally of any strengthening of anti‐female bias in the Chinese society), the pre‐existing level of anti‐female bias would, in any case, tend to make any restrictive policy fall disproportionately on the female child.

The restrictive developments of the post‐reform period include not merely the enforced control of family size, but also a considerable reduction in the general medical care and health services available in rural China. The ‘supported security’ on the basis of which China had achieved so much prior to the reforms has been weakened rather than strengthened in some important ways by the reforms affecting the economy and the society.45 In fact, despite the increase in outputs and incomes, the support system that the Chinese had built up with such success has been under severe strain. There is a clear weakening of commitment to public support measures, which may be partly ideological, related to the recent passion for economic liberalism.46 But it is also connected with the undermining of the financial and institutional basis of public support measures at the local level in rural areas, as a result of the abandonment of previous communal arrangements and their replacement by the ‘responsibility system’, which we discussed earlier.47

The rural production brigades used to offer a widespread cooperative medical system of health insurance, but the proportion offering this support has declined from 90 per cent in 1977 to about 34 per cent by 1985.48 It appears that the earlier cooperative health insurance survives only in 5 per cent of the villages, according to the Ministry of Public Health.49 The number of village‐level medical workers has fallen from 3.3 million in 1975 to 1.2 million by 1984. The number of ‘barefoot doctors’ working in Chinese villages and brigade (p.220) clinics went down from 1.8 million in 1977 to 1.3 million by 1984, and is known to have fallen greatly since. The number of female barefoot doctors fell even more sharply. By now, that entire system of medically imperfect but socially useful service is in complete decline, and the number of barefoot doctors is no longer reported in the Statistical Yearbook.50

Further, the balance between urban and rural areas in terms of medical services—less unequal in China than in most developing countries—has also been disrupted by the changes accompanying the reforms. By 1983 the three‐quarters of China's total population who happen to live in rural areas had to make do with the services of only half the practising doctors, i.e. a third of what the urban areas had in per capita terms.51

The impact of these declines in rural health services must have been significant in general and particularly detrimental to vulnerable groups, and its role in halting the rapidly declining trend of Chinese mortality would have been substantial. Also, the burden of decreased health services seems to have been unequally shared between boys and girls, and given the pre‐existing anti‐female bias (whether strengthened or not in the post‐reform period), the gender inequalities can be expected to be most consequential in periods of general contraction.52

Despite these disruptions of communal health facilities, China does, of course, remain considerably ahead of India in terms of its widespread public health provisions and related social security facilities. The Chinese also remain firmly in the lead in the fields of longevity, nutrition and health. But the recent economic reforms, with their negative effects on public support (especially at the local level in the rural areas), have moved China a little bit in the direction of India, and that—in this context—is not a particularly helpful development. The weakening of the support‐led system has not been outweighed by a growth‐mediated new development.

It is too early to judge what kind of a new equilibrium China will achieve. At the moment it is clear that there is a pause and perhaps even some regress in the expansion of longevity and in the reduction of morbidity, despite the progress in incomes and opulence. The authoritarian nature of Chinese politics has permitted an abrupt reduction in the social security provisions that had contributed so much to China's earlier successes.

It is not our purpose to prognosticate what will happen in China in the long run. From the point of view of public action against hunger and deprivation, what is especially important is an understanding of the policy issues raised by China's varied experience. A particularly important one concerns the role of public support, especially the universality of the coverage of public support. The contribution that universal (or near‐universal) support‐led security can (p.221) make to living conditions is exemplified both by (1) China's progress in expanding some of the most basic capabilities up to the late 1970s despite little increase in per‐capita GNP and food consumption, and (2) her comparative regress in these vital fields in the post‐reform period despite rapidly rising outputs and incomes.

11.5 China, India and Kerala

India offers within its own boundaries quite a variety of experiences. There are great interstate differences, and one state in particular, viz. Kerala, deserves special attention in terms of public action against hunger and deprivation. Kerala is one of the poorer Indian states.53 Yet it has achieved a remarkably high level of life expectancy—by a long margin higher than any other Indian state, including the richest states of Punjab and Haryana.

The last rigorous estimate of life expectancy at birth in India available at the time of writing this monograph is for the period 1976–80. The figures for the next half‐decade 1981–5 have not yet been published. The all‐India average figure in 1976–80 was 52 years, but that for Kerala was as high as 66 years (68 for females and 64 for males), which is not materially different from China's achievement.54 Early estimates indicate considerable increases in life expectancy in India as a whole, but Kerala's overwhelming lead has been maintained.

Doubts have been raised as to whether Kerala's outstanding longevity indicators are reflective of a comparable breakthrough in general health and nutritional well‐being.55 These doubts have arisen mainly from the additional evidence provided by (1) low calorie intakes, and 92) high self‐reported morbidity rates. The significance of these indicators is, however, open to serious question. As was discussed at some length in Chapter 3, calorie intake offers a poor basis of assessment of nutritional status, and when it comes to more direct measures of nutritional well‐being (especially the avoidance of severe undernourishment), it appears that Kerala remains firmly ahead of other Indian states.56 As far as high morbidity rates are concerned, these are (p.222) based on self‐reported illnesses, and it is not easy to determine the extent to which they reflect a greater level of articulation of a population that is enormously more literate and health‐conscious than people anywhere else in India. It has been argued, in fact, that self‐reported morbidity indicators as they exist today are extremely misleading in the context of interpersonal comparisons of well‐being.57 While there is some scope for disputation as to the precise relationship between longevity and other aspects of nutrition or health in Kerala, the overall picture of success is hard to deny.

The role of public support in Kerala's achievement has attracted justified attention.58 This has partly taken the form of extensive medical coverage of the population through public health services, helped by the determination of the population—much more educated than elsewhere in India—to seek medical attention.59 Kerala is also the only Indian state in which the public distribution (p.223) of food goes well beyond the limits of urban areas and provides significant support to the rural population.

The high literacy level of Kerala is also a major asset, especially in making people more eager and more skilled in seeking modern remedies for treatable ailments. It may also have a role in facilitating public participation in social change and in generating public demand for social security.60 The innovative programmes in the distribution of health care and food in Kerala have frequently followed articulated social and political demands.61 The same is true of a range of institutional changes, notably wage legislation and land reforms.

In contrast with the all‐India adult literacy rate of 41 per cent (age 15 or over) in 1981, Kerala's literacy rate is 78 per cent. That ratio is also substantially higher than China's 70 per cent for 1985. There is also a particularly important feature in Kerala's pattern of literacy. It shows relatively less gender bias. The percentage of adult literacy in Kerala in 1981 was 71 for women and 86 for men, compared with 26 and 55 respectively for India as a whole. The gender bias in literacy is also substantially less in Kerala than in China, as Table 11.4 indicates. The female literacy rate of 71 per cent in Kerala compares well with the corresponding rate of 56 per cent in China.

Table 11.4 China, India and Kerala: Selected Comparisons

Adult literacy rate (per cent)

Life expectancy at birth (years)

Female‐male ratio























(a) 1984 and 1985

(b) 1976–80 and 1981

Notes and sources: Adult literacy rates for China for 1985 are taken from UNICEF, The State of the World's Children 1988, and for India and Kerala, they are calculated from the data presented in the Census of India 1981. Life expectancy figures for China are taken from Banister (1987), Table 4.18, and relate to 1984. The official Statistical Yearbook of China 1986 gives figures of 69.3 years for females and 66.4 years for males for 1982. The life expectancy figures for India and Kerala relate to the period 1976–80, and are taken from SRS Based Abridged Life‐Tables 1976–80 (Occasional Paper No. 1, 1985), published by the Registrar General of India. The female–male ratio for China in 1985 is obtained from Statistical Yearbook of China 1986, and that for India and Kerala for 1981 from the Census of India 1981.

In fact, the history of literacy expansion through public action goes back a long time in Kerala. The state of Kerala was formed at the time of independence by amalgamating, on grounds of linguistic uniformity and cultural unity, two so‐called ‘native Indian states’ (viz. Travancore and Cochin) with a part of the old Madras Presidency in British India (viz. Malabar). As it happens, public policy in both the native Indian states put much greater emphasis on general education and literacy than was the case in the rest of India, and the emphasis on female education was particularly exceptional. In fact, as early as 1817, the ruler of Travancore, Rani (Queen) Gouri Parvathi Bai, had issued a rescript commanding that ‘the State should defray the entire cost of the education of its people in order that there might be no backwardness in the spread of enlightenment among them, that by diffusion of education they might become better subjects and public servants and that the reputation of the State might be advanced thereby’.62 The Rani was probably right, and the wide educational base in Kerala seems to have had a major impact on other public policies in that state (including medical care and food policy), in addition to encouraging intelligent health practice at the family level.

Gender bias is a topic of some interest in the context of both China and India. As was noted in Chapter 4, both China and India have an exceptionally low female–male ratio (FMR) in the population, and the issue of female survival is (p.224) an important one in both countries. Life expectancy at birth has been lower for females than for males in India until very recently and the cross‐over—bringing India in line with much of the rest of the world—is supposed to have occurred only in this decade.63 In China, the move in recent years has been in the opposite direction, and the female life expectancy, which used to be higher than the male, has now become lower (see Table 11.3 and the discussion in the last section). In both countries the female–male ratio remains dismally low —around 93 females per 100 males. It contrasts with the ratio of 1.05 or higher observed in those countries (e.g. in Europe or North America) in which there is little gender bias in health care and food distribution (though there may be much sexism in other areas), and it is much lower (as discussed in Chapter 4) than the sub‐Saharan African ratio of 1.02.

In contrast with both China and the rest of India, the female–male ratio in Kerala was higher than 1.03 in the last census, and is taken to have risen further since then. This is a higher female–male ratio than that for every region of the developing world for which we examined aggregated data in Chapter 4 (including sub‐Saharan Africa, which was the basis for our calculation of ‘missing women’).

To what extent the relative absence of gender bias in Kerala relates to its radical public policy (discussed earlier) is hard to say. It would be surprising if a greater level of female education—and less gender inequality in the sharing of education—had not contributed to better prospects of a plausible life for women, both through raising the status of women and through increasing female economic power and independence in ‘cooperative conflicts’ (discussed in Chapter 4).

But there are also other factors to be considered here, including the partially matrilineal system of inheritance in parts of Kerala and the relatively long history of its left‐wing activist politics. It would be well beyond the scope of the present inquiry to go into this important but difficult question, and the sorting out of different but interrelated causal influences can be a particularly hard exercise. But it is worth noting, as a preliminary observation for closer scrutiny, that the one state in India that has made extensive use of support‐led security has also been able to avoid some of the disastrous implications of gender bias that plagues so many parts of the world. The possibility of support‐led systems making a contribution to gender equality is something that would deserve further investigation.64

(p.225) The success of Kerala in achieving support‐led security adds force to the plausibility of following this route even when the economy is very poor. The fact that Kerala has achieved such success through careful and wide‐coverage public support shows how much can be achieved even at a low level of income, if public action is aimed at promoting people's basic entitlements and capabilities.

People's capability to conquer preventable illness and to escape premature mortality depends crucially on their command over basic necessities and their ability to use these with skill. Public support of education, health, employment, etc. can contribute both to that command and to the necessary abilities. The varying experiences of China and India, and the internal diversity of those experiences (both over time and over regions), bring out the importance of these roles in varying contexts.


(1) The World Development Report 1988 puts the Chinese life expectancy at birth in 1986 at 69 years. The Chinese official statistics based on the 1982 census place it at 68 years in 1981. Judith Banister (1987) gives a lower estimate based on corrections for incomplete reporting, viz. 65 years, for the same year.

(2) The World Development Report 1988 gives the figure of 57 years as life expectancy at birth in India in 1986. The last reliable estimate based on sample registration survey for 1976–80 puts the figure at 52.3 years, but later estimates also exist suggesting a life expectancy in the mid‐50s.

(3) See UNICEF (1988), Table 1.

(4) UNICEF (1988), Table 2. Serious doubts have, however, been expressed about the reliability of these figures of birth weights, for purposes of international comparison, especially since there are some variations in the criteria used.

(5) See World Bank (1983a, 1984a, 1985), Piazza (1986), Hussain (1987), Jamison and Piazza (1987), and Hussain and Feuchtwang (1988).

(6) UNICEF (1988), Table 4. The male and female adult literacy rates given by UNICEF for each country are respectively 82 and 56 in China and 57 and 29 in India. The total population averages have been obtained from these data by weighting the female and male figures by the ratio of females to males in the two countries.

(7) The picture is more diverse if intracountry differences are also considered. In particular, the public intervention programmes in the state of Kerala in India seem to have made the achievements of that state comparable to China's. The mixed nature of these contrasts also points towards some policy lessons. The issue is taken up in the last section of this chapter.

(8) China's lead in the field of literacy is also clear enough, though India has a much more extensive tertiary education sector with a much higher percentage of the relevant age‐cohort going to the universities.

(9) Wilfred Malenbaum (1956, 1959, 1982) has provided careful and illuminating comparisons of economic growth in the two countries over the decades. See also Swamy (1986a), and the previous literature cited there and in Malenbaum (1982). See also Perkins (1983, 1988) on methodological problems and substantive results in assessing China's growth performance.

(10) These results are not particularly sensitive to the exact years chosen for the backward extrapolation based on World Bank data. Putting together the GNP figures for 1982 and the growth rates for 1960–82, as given in World Development Report 1984, yields 1960 per‐head GNP figures of $106 for China and $196 for India, i.e. a Chinese GNP per head 46% lower than India's (on this see Sen 1985a: 78–80 and Table A.2).

(11) The poorest country covered in the World Development Report 1988 is Ethiopia with a GNP per head of $120 in 1986 (see Table 1, p. 222). This figure is much higher than the attributed GNP per head in China of $106 in 1965 (at 1986 prices).

(12) Kuznets (1966: 36).

(13) World Development Report 1988 itself shows the limitations of its estimates and draws attention to the assumptions on the basis of which the figures are arrived at (see pp. 290–1). The issue here is not one of accusing the World Bank of deception or perfidy, but to see the case for rejecting the comparative numbers that are presented in some of these widely used tables and which are in fact quoted again and again in discussions relating to China and India.

(14) On the rapid revisions of China's GNP figures made by the World Bank, see Malenbaum (1982): ‘As noted, Bank data for China's GNP per capita were of the order of $400 for 1976 and 1977 (current US prices). Comparable figures for India were $150, some 37.5% of the Chinese levels. In 1980, the respective figures for 1978 were listed as $230 and $180 (1978 US prices); the ratio was 78.5%. The levels for 1979 are now shown as $260 and $190. A comparative development history that pictured China with a long‐period growth advantage of about 150% per capita over India by the late 1970s now shows an advantage only one‐quarter as large, some 30%–35%, over the same period. No structural explanation is offered for this radical revision’ (pp. 81–2). To continue the story where Malenbaum left off, the World Bank made the figures converge for all practical purposes by 1986, with (as mentioned in the text) a GNP per head of $300 for China and $290 for India. While the GNP figures for China were made to come down closer to India's GNP per head, the figures for GNP growth rates of China were kept sharply higher than those of India. What we are looking at here is not an account with just some little inconsistency somewhere, but one that is altogether contrary within its own internal structure.

(15) See Riskin (1986, 1987).

(16) See Banister (1987: 354). See also Carl Riskin's (1986) paper on the strategy of ‘feeding China’, and the empirical studies cited there.

(17) See World Bank (1983a, 1985), Jamison et al.] (1984), Perkins and Yusuf (1984), L. C. Chen (1988), and Shao (1988), among many other contributions.

(18) World Development Report 1988, Table 29, p. 278. Population per physician is reported to be 3,700 in India and 1,730 for China in 1981. The corresponding figures for population per nursing person are 4,670 in India and 1,670 in China. It is often claimed, with some justice, that the medical training programmes in India are typically more exacting than the Chinese programmes have been (at least until recently). But the big difference in having some medical services available across the country (even in the remotest parts) makes the Chinese overall situation altogether distinct from that in India, and in this achievement the logistic advantages of having many more doctors and nurses per unit of population have been quite crucial.

(19) For some interesting comparisons between China and India in the domain of health, see A. Bhalla (1987); see also Chen (1987). The Chinese arrangements, especially for the rural areas, provide a major contrast with the lack of penetration of health services in rural India, especially among vulnerable groups (see Bose and Tyagi 1983, and Murthy et al. 1988).

(20) On the nature and reach of the social security system and public services in China, see Riskin (1986, 1987), Ahmad and Hussain (1988) and Hussain et al. (1989), and the literature cited there.

(21) Various regions of China also suffered from adverse weather conditions during 1959–61, involving drought and flooding in different parts of the country. These conditions certainly contributed to the production problems, but it is hard to escape the conclusion that the bulk of the problem was caused by the failure of the policy changes initiated by the Great Leap Forward. In fact, the adverse climatic conditions were particularly important in making it harder to identify precisely how fully the ‘Leap’ had failed.

(22) On this see Sen (1982c, 1983a), Bernstein (1983), Riskin (1986, 1987).

(23) See Coale (1981), Aird (1982), Ashton et al. (1984), Peng (1987). The largest estimate of 29.5 million (Ashton et al. 1984) attempts to take full note of unreported deaths. Peng's (1987) estimation based on provincial demographic data suggests a figure of 23 million.

(24) That figure was obtained also by using the same method of comparing actual mortality with pre‐famine mortality, and making allowances for unreported deaths (Sen 1981a, Appendix D). The official estimate of excess mortality was 1.5 million, but that involved an undercounting due to incomplete temporal coverage.

(25) It must, however, be remembered that since the Chinese mortality rates had come down sharply already prior to the famine, the ‘extra death’ estimates based on pre‐famine mortality rates are in comparison with a pre‐famine death rate lower than that of most poor countries in the world. But even if considerably higher pre‐famine mortality rates are used, the excess mortality in China still amounts to astonishingly high figures.

(26) See Peng (1987: 653–5, esp. Table 3).

(27) See Riskin (1986, 1987) and Peng (1987).

(28) The data are taken from Peng (1987).

(29) On this see Peng (1987). ‘Ironically,’ notes Peng, ‘almost all the provinces that were praised by a People's Daily article for their “good performance” in establishing rural mess halls experienced severe excess mortality’ (p. 664).

(30) The Governor of Bengal, Sir T. Rutherford, wrote to the Viceroy of India explaining that a famine had not been declared to avoid the obligation to undertake the relief measures mandated by the Famine Codes. See Mansergh (1973, 363, Document No. 158).

(31) The reasons for this diagnosis and the empirical evidence for this view are discussed in Sen (1982c, 1983a).

(32) Bernstein (1984: 13).

(33) The figures are from the Statistical Yearbook of China 1981. See also Riskin (1986, 1987) and Jowett (1988).

(34) Riskin (1987), Table 6.5. See also Bernstein (1984) and Peng (1987).

(35) In 1962, shortly after the famine, when the recent experiences were being reviewed, Mao Zedong noted the problem of informational failure for planning in the absence of local democracy: ‘If there is no democracy and ideas are not coming from the masses, it is impossible to establish a good line, good general and specific policies and methods . . . Without democracy you have no understanding of what is happening down below; the situation will be unclear; you will be unable to collect sufficient opinions from all sides; there can be no communication between top and bottom; top‐level organs of leadership will depend on one‐sided and incorrect material to decide issues, thus you will find it difficult to avoid being subjectivist; it will be impossible to achieve unity of understanding and unity of action, and impossible to achieve true centralism’ (Mao Zedong 1974: 164). On this pronouncement and its context and relevance, see Sen (1983a, 1984a).

(36) On this see Sen (1982c, 1983a, 1984a) and Ram (1986).

(37) This is so with the highest of the estimated mortality figures, viz. 29.5 million (due to Ashton et al. 1984). If instead we take, say, Peng's figure of 23 million, then every six years there is more extra mortality in India than in the Chinese famine of 1958–61.

(38) See Riskin (1986, 1987), Johnson (1988) and Perkins (1988). On the incentive problems of collective agriculture, see Putterman (1986), and also Nee (1986).

(39) See Riskin (1988: 21).

(40) See Banister (1987) and Hussain and Stern (1988).

(41) Hussain and Stern (1988) argue that ‘the year to year fluctuations [in death rates] are likely to be largely the result of changing data sources and methods’, but confirm a broad pattern of ‘a reduction [in the death rate] to the end of 1970s followed by an increase of around 7% in the death rate from 1979–86’ (p. 18).

(42) See Banister (1987: Chapter 7). ‘Many couples, determined to have a son, have killed their infant daughters, either outright or by severe neglect, so that they could try again for a son’ (Banister 1987: 40).

(43) See Hull (1988) on this general question, and also on the results of a 1 per cent survey of the population carried out by the State Statistical Bureau in 1987.

(44) On this see Aslanbeigui and Summerfield (1989). On the comparison with the pre‐reform situation, see also Croll (1983), Hemmel and Sindbjerg (1984), Nee (1986), Wolf (1987) and Kelkar (1989).

(45) It is, in fact, also possible that the breakdown of social support provisions regarding old age security may have added to the existing ‘pro‐male‐child’ bias, given the association of male‐preference with the motive of insuring support in old age. On related matters, see Hussain et al. (1989).

(46) Riskin (1988) notes that the style is a hybrid one involving ‘a strange mixture of residual socialist rhetoric and Chicago School values’. But the socialist rhetoric has been distinctly in retreat in recent years.

(47) On this see Banister (1986, 1987), Nee (1986), Panikar (1986), Bhalla (1987), Ahmad and Hussain (1988), Johnson (1988), Shao (1988), and Hussain et al. (1989).

(48) See Banister (1986: 2–3).

(49) See Shao (1988).

(50) From 1986 the Statistical Yearbooks discontinued giving the numbers of barefoot doctors, and the last year of report is 1984. See also World Bank (1984a) and Banister (1986).

(51) See Banister (1986: 2).

(52) On the general issue of gender bias and intrahousehold distribution, see Chapter 4.

(53) The standard comparative data can be found in Agrawal, Verma and Gupta (1987). For a probing analysis of the comparative real incomes of the different Indian states (taking note of distributional differences), see Bhattacharya, Chatterjee and Pal (1988). See also Sanyal (1988) on Kerala's comparative poverty in terms of landholding.

(54) This figure of 65.5 years is, in fact, a little higher than Judith Banister's estimate of life expectancy for China even before the reforms (65.1), and a fortiori so afterwards (64.6 in 1984 according to Banister's estimate). It must, of course, be mentioned that there is a lot of regional diversity within China (on this see Prescott and Jamison 1985), and the life expectancies in the more advanced regions are higher than that in Kerala.

(55) See e.g. Panikar and Soman (1984) and B. G. Kumar (1987).

(56) See e.g. the anthropometric evidence presented in B. G. Kumar (1987), Table 6.7, and Subbarao (1989), Table 6 (both based on survey data collected by the National Nutrition Monitoring Bureau). The very low incidence of ‘severe undernutrition’ in Kerala is particularly striking—a matter of particular importance for health, well‐being and survival (on this see Lipton 1983, and also Chapter 3 above). The percentage of ‘severe undernutrition’ for children between 1 and 5 years of age in 1982 was 6.1 for India as a whole and only 1.5 for Kerala (Kumar 1987, Table 6.7).

(57) For the evidential basis of this claim, see Murray and Chen (1989). As shown by these authors, many contrary findings arise from assessments of well‐being based on self‐reported morbidity data. For instance, while self‐reported morbidity rates in India are highest in Kerala (and lowest in Uttar Pradesh, where the expectation of life is about 25 years shorter than in Kerala), they are even higher in the United States, and within the United States they are highest in the higher income groups. Such findings do not imply, of course, that self‐reported morbidity data are entirely useless. Self‐perception of morbidity as well as articulate reporting are of obvious interest in the social analysis of perception and communication.

(58) See Centre for Development Studies (1975), S. K. Kumar (1979), Gwatkin (1979), Gwatkin et al. (1980), Jose (1984), Panikar and Soman (1984), Halstead et al. (1985), Krishnan (1985, 1989), Stewart (1985), Caldwell (1986), B. G. Kumar (1987), Osmani (1988a).

(59) It is arguable that the higher incidence of reported morbidity can be seen as a positively contributory factor in Kerala's success in dealing with diseases and in expanding longevity. Treatment can begin only when attention is paid to the disease from which one is suffering. A large proportion of the Indian rural population die from undiagnosed—often ignored—diseases.

(60) See Halstead et al. (1985) and Caldwell (1986). See also Mencher (1980).

(61) The radicalism of Kerala's politics found some expression in 1957 in its being the first Indian state to elect a communist government. Since then different political coalitions have ruled the state, and the parties in office have been typically kept on a short leash.

(62) Quoted in Census of India, 1931, vol. xxviii (Trivandrum, 1932), p. 301.

(63) See Dyson (1987).

(64) It should be mentioned here that the female–male ratio being greater than unity has been a feature of Kerala since before this century. In the Census of India for 1881, it was remarked: ‘In Travancore, as in other southern populations, the proportions of the sexes approach more nearly to European standards than is the case in the northern states and Provinces’ (Report on the Census of British India, vol. i (London: HMSO 1883), p. 70). This fact should not be seen as automatically turning the table against the importance of support‐led security for reducing gender bias, since the active promotion of primary education, especially female education, in Kerala (in particular in Travancore and Cochin) goes back to early in the last century. Both the features of higher FMR and higher involvement in basic education (especially for girls) go back to the 19th century in the case of Kerala.