The four papers in this special issue of China Economic Review constitute a major contribution to the understanding of China's Great Leap Famine. There is little doubt that this famine was the worst of modern times in terms of the human suffering and excess deaths. Yet there has been surprisingly little systematic analysis of the factors responsible for the famine. We are indebted to the authors of these papers for adding substantially to our knowledge concerning the famine. Due to the reluctance of the Chinese government to permit or encourage public discussion and research on the causes of the famine or on its effects on the Chinese people, the Great Leap Famine is the least studied and understood of the major famines that have occurred in the world.
    As you will see if you read each of the papers, there is a lack of unanimity on several important points. This is understandable when an event of such enormous magnitude is under analysis, an event that occurred within an exceedingly complex setting with many different actors, and with much that remains unknown concerning what actually happened.
    It is not my objective in these introductory remarks to take sides where there are evident disagreements, as there are. It would be foolish of me to do so since three of the seven authors are my former students and another is a faculty colleague. What I shall do is to briefly highlight what I believe are the major points of agreement, then turn to the important points of disagreement or expressions of different viewpoints concerning some aspect of the famine, and conclude with what I consider to be a small number of unanswered questions.

Points of agreement
    1. The number of excess deaths was enormous though on the basis of available records it is not possible to provide an estimate that would be accepted by a group of unbiased analysts. But there can be little doubt that the death toll was enormous. Whether the excess deaths numbered 20 million or 30 million, as a number of qualified studies indicate, the enormity of the famine is evident. Obviously the lives of millions more were severely affected by hunger and famine-related diseases.
    2. The exaggeration of the 1958 grain output estimates clearly affected the course of the famine by misleading high officials, local officials and individual farm families into believing that normal approaches to conserving food were unnecessary. The political and ideological climate that made an estimate of 375 million tons of grain believable to high governmental and party officials - an amount double the 1957 output - can only raise questions about where rationality had gone. The belief that the country was awash in grain clearly influenced policies, such as the amount of procurement, the continued exportation of grain and the institution of the free supply system for grain, as well as enhancing beliefs in the glorious achievements of the commune system and other aspects of communism. Obviously if the country had produced twice as much grain as the year before, there was little reason to economize in the use of the available supply.
    3. A remarkable and distinguishing characteristic of the famine is that it began following the largest grain crop on record and prior to the spring harvest of the subsequent year. There is disagreement, of course, of the factors causing the excess deaths that occurred in late 1958 and in the early months of 1959. While Lin & Yang disagree with the role attributed to the communal kitchens by Wen & Chang in causing famine deaths in late 1958, there is agreement concerning the fact that, whatever the reasons, there were deaths that could be attributable to famine before the 1959 spring harvest. Lin & Yang note the large increase in mortality in 1958 over 1957 in Sichuan, Yunan and Gansu - the increases in death rates ranged from 37 percent to 120 percent. The famine deaths were attributed to the large exports from the provinces to the rest of China and by implementation of many of the labor intensive projects of the Great Leap that resulted in excess consumption due to the energy requirements of the heavy work. As of 1957, the nation's calorie supply was generally adequate for light to moderate work most of the year and for only short periods of intense activity. But whatever the reason for the famine conditions in these three provinces, there is agreement that the famine started in 1958 - following a record grain crop. This makes the Great Leap Famine unique among the major famines of modern history. Even Stalin's punitive famine against the Ukrainians followed a damaged harvest.
    4. The severity of the famine was the result of many variables, including excessive procurement, the protection of urban consumers from the worst effects of the food shortages, the continued export of grain while people starved, the output declines in 1959, 1960 and 1961, and waste of food in the communal dining halls.
    5. The famine was primarily induced by policies adopted during the Great Leap Period. While, as noted, grain output declined in 1959, it is argued that the decline was largely policy induced. It was policy induced in at least two senses. First, following his acceptance of the inflated 1958 grain production figures, Mao argued that the area devoted to grain should be reduced and the actual area in 1959 was 10 percent below that of 1958. Second, there were a whole series of radical policies adopted that reduced production incentives. One of these, as argued by Lin, was that farmers were denied the right of exit from the communes and this had significant negative effects on their willingness to work. The 1959 output decline was not serious enough to have caused a large number of famine deaths if the supplies of food were reasonably well shared among the population and if the available food supply had been managed by individual households. The per capita level of production was substantially greater than in 1962, the year the famine ended, and greater even than in 1963 when mortality was lower than in 1959. While policies affected the output levels, how the available output was shared among rural areas, between rural and urban areas, between China and the rest of the world and how well the existing supplies were managed in consumption almost certainly had a greater impact on famine deaths.
    6. The famine ended in 1962 even though the per capita availability of grain was well below that of 1959.

Areas of disagreement
    There are, of course, several areas of disagreement. The major area of disagreement relates to the differences concerning the effects of the various policies of the Chinese government. Two points may be noted. The first is the attribution of responsibility for the famine starting in 1958 and early 1959, following a bumper harvest. Wen & Chang argue that this was due to the excessive consumption and waste of food engendered by the communal dining system while Lin & Yang argue that the famine in 1958 was concentrated in three provinces and was primarily due to excessive exports to the rest of China, engendered by over enthusiastic provincial party officials seeking favor from Beijing. Yang & Lu also date the famine from 1958 but do not dwell on the factors responsible for the early start.
    The second is that of the factors responsible for the severity of the famine. Lin & Yang argue that a combination of output decline caused by various policy factors combined with the favoritism of urban residents over rural residents and the continued export of grain were the primary causes of the severity of the famine. Wen & Chang argue for the excessive consumption and waste of food associated with the communal kitchens, combined with the output reduction, as the major cause of the severity. Yang & Lu emphasize the political factors associated with the spread of the communal kitchens, combined with the radicalism of provincial leaders, for the differential impacts of the famine. Yang & Lu make the point that the free supply system and the communal kitchens created a problem of the commons; there was no incentive for any individual family to conserve on food consumption since to do so would not increase the supply available to them in the future.

Unanswered questions
    The major unanswered question - and one that probably can never be answered - is the determination of the joint and independent effects of the various factors that resulted in the famine. There were so many policy factors that mediated between the available food supply and its distribution among the people as well as the policy factors that influenced the level of production of food that it would require a very complex model and enormous quantities of data to permit a systematic study. As Riskin notes in his informative discussion of seven questions, the state statistical system was in great disarray - and that statement may in fact exaggerate the actual state of affairs.
    While the available information on the number of famine deaths is more than adequate to portray the horror that descended on the rural people of China, it is not obvious that the data on mortality by province are sufficiently accurate to support a detailed analysis of the effects of the various relevant variables on mortality. Nor can one be sure that either provincial or national production data are sufficiently accurate to support a sophisticated statistical analysis. Nor are there data on state procurement for all provinces and thus it is not possible to determine with a reasonable degree of accuracy the per capita food supplies by provinces. It was not only that the statistical services, as Riskin rightly notes, were disrupted by the frenzy of the Great Leap Period but there was hardly time for recovery before the beginning of the Cultural Revolution with its enormous disruption of most governmental functions, including that of the statistical services. Much of the available data relevant to the understanding of the famine were assembled in the late 1970s and early 1980s, by reports from local areas, such as villages, townships and counties. What is not clear is how much of the evidence was derived from actual records and how much was based on recall.
    The most likely repository of relevant morality data for the period are from the village records. I have seen village records that give population data for the period of the famine. Presumably detailed, though not necessarily accurate, data on production and inputs would be available from the commune, brigade and production team records. However, it is highly probable that most of the commune records have long since been lost or destroyed since these records belonged to the accountants and they were required to keep them for only five years. Apparently only a few commune accounts have been collected and would be available for research.
    One of the most striking deficiencies in the available data is the movement of grain from province to province. Production data, even if fully reliable, are not sufficient to understand the role of inadequate supplies as a cause of the famine. It matters little whether the cause of inadequate supplies is a decline in output, an excessive export of grain from the province or the giving of preference to urban consumers at the expense of farm people. In the case of the Great Leap Famine, the food availability decline (FAD) for a province could be due to either or both a decline in production or an increase in exports from a province or any other relevant geographic area. It may be possible to explain the declines in national production by policy changes, such as the abolition of the right of exit, but much more needs to be known to explain differences in output response by provinces. But even more challenging would be the estimation of the factors associated with differential provincial exports or in understanding why China continued to export grain while officials knew that some of its people were starving. Given the availability of published data, it is not possible at this time to reconstruct the provincial supply situation during the period of the famine.
    Even if we had data on the net provincial supply of grain and other foods, would it be possible to explain the behavior of the responsible officials with regard to the volume exports from the various provinces? Can we ever understand why some provincial officials gave priority to limiting the amount of grain that left a province while others seemed more inclined to provide the maximum amount in response to actual or implied demands from Beijing?
    Finally, why did the famine end in 1962?¹ The overall supply of food, at least based on the official data, was clearly less than in 1959 and very nearly the same as in 1960, the worst year of the famine. The per capita grain consumption in 1960 has been officially estimated at 164 kilograms and at 165 kilograms in 1962. There were many policy changes that were instituted, starting in 1961 or perhaps even earlier. These included the abolition of the community kitchens, the institution of private plots, the allocation of land to households in some areas, the significant reduction in the percentage of grain output that was procured, made possible at least in part by a significant level of grain imports.
    The impact of the change in China's net trade position in grain in 1960, when net exports were 2.7 million tons, and in 1962, when net cereal imports were 4 million tons, was larger than one might at first thought believe it to be. The quantities are in terms of trade grain. The difference in national availability of grain represented by the shift in net trade would have provided 120 million people with an average of 500 calories per day for the year. If this supply had been concentrated in the eight provinces with the highest rates of mortality, there can be little doubt that the severity of the famine would have been reduced by an order of magnitude. The total population of the eight provinces in 1957 was 244 million. An increase of 500 calories per day for one person out of every two would have made a significant difference in the number of famine deaths in 1960.

Could it ever happen again?
    Obviously one cannot be certain that China will never again be subjected to a major famine, even one far less serious than that of the Great Leap period. But if a famine that approached that of the Great Leap period were to occur, it would require policy errors of even greater magnitude than those that occurred thirty years ago, if one can imagine that such would be possible.
    The basic conditions with respect to the supply of food in China today are radically different than they were four decades ago. First, the daily per capita food supply has increased from less than 2,300 to well over 2,600. This is a large difference. As noted earlier, an average intake of 2,300 per capita did not provide enough for adult workers to engage in strenuous physical activity throughout the year - food needed to be reserved for periods of heavy activity, such as during plowing, planting and harvesting. At the current level of consumption, there is a significant margin of safety. In addition, an increased percentage of the population now does little physical work and has relatively low calory requirements.
    Second, the stocks of food in the possession of the farm population are now very large, while in 1958 there was a deliberate campaign to take from farm people any food stocks that they might have had. While subject to some dispute, the grain stocks held by farmers are now very large, amounting to several hundred million tons at the end of the calendar year. But the situation with respect to other stocks of food also differ substantially. Many farmers slaughtered their farm animals rather than turn them over to the communes in 1958. The number of hogs per capita of the national population is substantially greater than during the 1950s and much greater than the number in 1958. These and other farm animals represent a substantial stock of food that could and would be drawn upon in case of a severe food production shortfall.
    Third, China is now much more open to the world than it was in 1958. It could draw on large quantities of food from the rest of the world - it could without significant difficulty fully offset a 7 or 8 percent shortfall in grain production through importing as much as 40 million tons of grain. But it would not be necessary to fully offset the grain production shortfall. In any case, a grain production shortfall in a given year of 10 percent would not result in famine in China today. Such a shortfall from the current level of 500 million tons would only take grain output back to the level of 1994, a year in which China was a modest net exporter of grain.²
    Fourth, China now has a much more active communication system than it had in 1958. One of the factors that contributed to the severity of the Great Leap Famine was the control over the flow of information by the Communist Party. Today the press is much freer and there are many more sources of information available. Private or personal communication is now widely available through the telephone and, increasingly, e-mail. There are now millions of rural residents temporarily residing in cities that have ready access to information about conditions in rural areas. Even if a government wanted to control access to information about a major famine, if it should occur, it would be impossible to do so. Compared to four decades ago, there is now much greater openness concerning natural disasters. In the past two decades there have been open references to famine conditions and to problems faced by people confronted with other natural disasters such as floods.
    Finally, in spite of a recent effort of the government to suppress the grain market,³ most food products are available to consumers through markets where prices respond to supply and demand conditions. In the period of the Great Leap Famine nearly all market activity was restricted and there were few or no price signals of the famine as it spread. Today, and in the future, famine conditions would be immediately reflected in the market prices for the majority of food products. While grain continues to provide a major source of all calories in China, it accounts for only 7 percent of consumption expenditures for urban residents and, perhaps, for 15 percent for rural residents.  Thus conditions of food stringency would be evident in the market prices for most food consumed by both urban and rural residents. Only if there were significant further retrogression from reforms leading to a socialist market economy, would markets not provide a prompt signal of impending famine conditions.

Concluding Comments
    The four papers have significantly increased our understanding of the Great Leap Famine. We are indebted to the authors for their efforts to help us understand the factors that resulted in such an enormous human tragedy. Let us hope that they continue their scholarship on this topic and add further to our knowledge in the future.

1.  In some of the worst affected provinces, such as Anhui, Gansu and Qinghai, the famine was over in 1961 when mortality had returned to approximately the same levels as in the years prior to 1958.

2.  China would need to change the manner in which it manages international trade in grain if it were to make effective use of trade to offset a domestic production shortfall. The reform process seems not to have affected how trade in grain is managed. It is still handled as though the world market in grain doesn't exist and operates on the assumption that the volume of trade needs to be determined no later than the beginning of the calendar year and contracts made at that time. It is not obvious why the calendar year is relevant - this means that trade volumes for 1996 were made before there was any reliable knowledge of the 1996 grain crop, which as a bumper crop, nearly 8 percent larger than in 1995, which had also been the largest crop on record up to that time. In spite of these bumper crops, China imported substantial amounts of grain in both 1995 and 1996, imports that could in no way be justified by the domestic supply situation. In 1994 modest quantities of grain were exported even though the nominal price of grain increased by approximately 50 percent.
    The effect of the bumper grain harvest in 1996 combined with the prospects for a good harvest in 1997 resulted in a substantial decline in grain prices and an unsuccessful effort to provide a price support to increase the returns to farmers.
    It is foolish to plan trade in grain prior to knowledge of the size of the grain crop and it is equally foolish to make contracts for the entire year.

3.  In the name of reform, it was announced in People's Daily (May 26, 1998) that it was the intention of the Government of China to monopolize the nation's trade in grain. It has become illegal for private traders to purchase grain from farmers and as soon as private traders exhaust their stocks, all transactions in grain will be controlled by governmental agencies. The objective of the reform is to increase the price of grain received by farmers, apparently because it is feared that the current low prices of grain will discourage production in the future. Contrary to projections made by certain alarmists, such as Lester Brown, China now "suffers" from excessive supplies of grain due to high production levels and large carryover stocks. The real objective of the reform appears to be to prevent the government from suffering losses due to the higher prices paid to farmers - the higher costs are to be passed on to the consumers. In the past efforts to increase prices paid to consumers have resulted in large losses to the government. In the short run, the reason for the large losses was that the government failed to buy enough grain to raise market prices to the price support level. It mistakenly blamed its losses on the action of private traders, who bought grain from farmers at less than the support price and sold that grain into the market. Since the available supply is greater than the demand at the support price, the government must buy enough grain to limit the market supply to the amount that will be demanded at the support price. This means that it must buy a larger quantity than will be sold into the market and must add to already very large stocks. If it buys enough grain, it would make no difference whether private traders were permitted to remain in the market.