2. La crisis de la escuela moderna y sus rituales: ¿hay lugar para la memoria?
3.3. La representación del pasado en los jóvenes
With the relatively strong harvest of 1998 and the upturn in foreign assistance, the worst of the famine was probably over by the end of the year. How was the misery of the famine distributed? Who had been most seriously affected?
One interpretation is that the famine was a classic food availability fam- ine and that the government did all in its power to distribute food as equally as it could (S. Lee 2003, 2005; Woo-Cumings 2002). In such a setting, the command-and-control features of the socialist system might even have been an advantage because of the power of the state to command and reallocate resources. Under this interpretation, while a very small core elite might have been shielded altogether from the famine’s effects, the rest of the population shared relatively equally in the declining food that was available.
Three types of data allow us to test this claim, at least for 1997–98: variation in provincial production of grain, variation in farmers’ rations by province, and variation in provincial distribution through the PDS. These data show that while production varied enormously across provinces, the government main- tained a relatively common target for farmers’ rations in 1997. If we consider the problems that the government faced procuring grain following the 1994–96 harvests, then the effort to raise farmers’ rations and to keep them constant across provinces was almost certainly a conscious effort to limit diversion.
A consideration of the PDS data in table 3.5, however, suggests anything but equality across provinces. Using data supplied by the government on the supply of food to the PDS population by provinces and the share of the population dependent on the PDS, we can calculate the per person allocation of grain across provinces during the last part of the famine, from September 1997 through April 1999.19 The differences are striking. Pyongyang consistently comes out on top, sometimes receiving per person rations that are nearly twice those in less protected provinces. Chagang and the rice bowl of South Hwanghae also see higher alloca- tions, while two cities—Namp’o and Kaeso˘ng—and the northwest and northeast provinces fare much less well. The government appears to do a quite remarkable job of allocating grain after the harvest of 1998; indeed, the uniformity of distribu- tion is even suspicious. But by that point, the worst of the famine had passed.
TABLE 3.5. Monthly PDS Allocations, November 1997–April 1999 (kg per person)
Nov. 97 Dec. 97 Jan. 98 Feb. 98 Mar Apr– Aug 98
Sep-98 Oct. 98 Nov. 98 Dec. 98 Jan. 99 Feb. 99 Mar. 99 Apr. 99
Pyongyang 9.9 9.9 7.4 4.9 1.0 0.0 3.7 3.7 8.5 8.5 5.6 5.6 4.2 0.9 S. Phyongan 6.6 6.6 4.9 3.3 0.7 0.0 2.5 2.5 8.6 8.6 0.6 0.6 4.3 0.9 N. Phyongan 6.4 6.4 4.8 3.2 0.6 0.0 1.6 1.6 9.1 9.1 6.0 6.0 4.4 1.0 Chagang 10.5 10.5 7.9 5.2 1.0 0.0 2.6 2.6 8.7 8.7 5.7 5.7 4.3 0.9 S. Hwanghae 8.8 8.8 6.6 4.4 0.9 0.0 3.3 3.3 8.2 8.2 5.5 5.5 4.1 0.9 N. Hwanghae 8.8 8.8 5.3 3.6 0.7 0.0 2.3 2.3 8.0 8.0 5.3 5.3 3.9 0.8 Kangwo˘n 5.6 5.6 4.1 2.8 0.5 0.0 1.4 1.4 7.9 7.9 5.3 5.3 4.0 0.9 S. Hamgyn 6.7 6.7 5.0 3.4 0.6 0.0 1.3 1.3 8.5 8.5 5.7 5.7 4.3 0.9 N. Hamgyo˘n 7.5 7.5 5.8 3.9 0.7 0.0 1.5 1.5 8.5 8.5 5.6 5.6 4.2 0.9 Ryanggang 9.5 9.5 7.0 4.7 0.9 0.0 1.9 1.9 8.3 8.3 5.6 5.6 4.1 0.9 Kaeso˘ng 8.2 8.2 6.0 4.0 0.8 0.0 1.8 1.8 7.2 7.2 4.8 4.8 3.6 0.8 Namp’o 6.8 6.8 5.1 3.4 0.6 0.0 1.7 1.7 8.2 8.2 5.5 5.5 4.0 0.9 Total 7.9 7.9 5.8 3.9 0.7 0.0 2.3 2.3 8.4 8.4 4.9 4.9 4.2 0.9
Note: Calculated as monthly PDS allocation to province as a share of the province’s PDS-dependent population.
Sources: November 1997–October1998: FAO/WFP (1998b, table 6); November 1998—April 1999: FAO/WFP (1999a, table 3); Population by province: FAO/WFP (1999b, table 6) 8/ 23/ 06 1 0: 57: 1 7 A M
70 PERSPECTIVES ON THE FAMINE
Further insight into regional differences emerges from refugee interviews. Starting in September 1997 and continuing over a period of more than a year, the Korean Buddhist Sharing Movement (later Good Friends) researchers interviewed nearly two thousand North Korean refugees in the Chinese border area. These interviews combined more open-ended testimonials with structured questions designed to document the rise in mortality, the decline in birth rates, and the coping behavior of households (see, e.g., KBSM 1998; Good Friends 1998, 2004). As the Good Friends researchers freely admitted, the interview- ees were not randomly selected and overrepresented both geographical regions close to the Chinese border, particularly the northeast, and households that were the most vulnerable relative to the nation as a whole. Nonetheless, with the appropriate cautions and adjustments, the Good Friends work tells us a tremendous amount about the famine.
An update of earlier research by Good Friends (1998) based on a sample of 1,694 refugees asks which areas they believe were the hardest hit. Sixty-two percent said South Hamgyo˘ng, 23 percent said North Hamgyo˘ng, and 22 per- cent said other provinces; only 9 percent said that provinces were experiencing equal levels of distress. These responses gain some credence because they do not simply mirror the residences of the refugees. Nearly 60 percent of respondents came from North Hamgyo˘ng and only 20 percent from South Hamgyo˘ng; nonetheless, a substantial majority saw the latter province as the more seri- ously affected. Information on mortality provides further confirmation of these regional differences. The average mortality of all respondents’ families in the sample is 28.7 percent, but this ranges from a low of 16.7 percent for Pyongyang to 32.1 percent for South Hamgyo˘ng and 32.9 percent for Chagang.
A plausible reason for this difference between the two provinces is that it was somewhat easier for the residents of North Hamgyo˘ng to move across the bor- der or to benefit from black-market exchanges and trade. These responses also provide more circumstantial support for Natsios’s claim with respect to triage, particularly when we consider that food aid did not start to flow directly into the northeast until the second half of 1997. As we discuss in chapter 7, UN- sponsored nutritional surveys provide striking evidence of regional disparities in nutritional status and modest support for the notion that South Hamgyo˘ng was disadvantaged relative to North Hamgyo˘ng.
Both the nature of the PDS and the interview evidence suggest that the most severely affected were urban households in the disadvantaged provinces.20 Without any direct claim on the harvest, with no access to private sources of supply, and with inadequate money wages to command food through the relatively limited market channels, urban workers and their families were com-
The Distribution of Misery 71
pletely at the mercy of a faltering PDS. This finding also gets support from ref- ugee interviews. In the 1998 Good Friends study just cited, the overwhelming majority—88.7 percent of those answering—said that urban areas were more severely affected than rural ones. Only 9.5 percent said that urban and rural areas fared the same, and only a handful of respondents—1.8 percent—said that rural residents did worse.
The occupational data on mortality from the 1998 survey also provides a number of interesting clues about the breakdown of the urban industrial econ- omy (table 3.6).21 As we would expect, office workers, professionals, and sol- diers—in descending order of vulnerability—were more protected than other occupational groups. Although manual laborers constitute the largest group of family members, they also show somewhat lower than average mortality rates. What is striking is the large number of family members who are identified as jobless. While this could be capturing the elderly, refugee testimonials sug- gest that factories in the major industrial cities were effectively left idle by the absence of inputs and energy. New investment, and thus construction, had also ground to a complete halt. Despite the guarantee of employment that the socialist economy presumably provides, the industrial sector was almost certainly going through a process of informalization quite similar to that seen in other developing countries during economic crises. Some strata of more marginal workers—perhaps in sectors such as construction, in less favored cit- ies and industries—were undoubtedly the very hardest hit.
TABLE 3.6. Mortality Rate by Occupation
Occupation Family Members Mortality Mortality Rate
Manual labor 2,398 441 18.4 Office worker 633 75 11.8 Professional 43 3 7.0 Farmer 296 71 24.0 Student 1,951 336 17.2 Soldier 217 13 6.0 Housekeeper 284 95 33.5 Other 122 27 22.1 Jobless 1,769 807 45.6 Unknown 1,536 785 51.1 Total 9,249 2,653 28.7 Source: WFProgramme 1998.
72 PERSPECTIVES ON THE FAMINE
What about the rural sector? It is revealing that only a small share of the refugees interviewed—3.2 percent—were farmers, but whether this reflected tighter control at the village level and greater difficulty of movement or better material circumstances is difficult to say. Although the share of farmers in the Good Friends sample is small, the reported mortality rate among farm families is only slightly lower than the mean (24 percent vs. 28 percent). Even if in prin- ciple the allocation granted to farmers was relatively even across the country, the weather resulted in regional and more localized production shortfalls that pushed cooperative farm supplies near or below subsistence thresholds.
Thus, while farmers as a whole undoubtedly did better than urban dwellers, residence mattered: farmers living in areas strongly affected by the floods and drought depended largely on international largesse and the capacity of the cen- tral government to reallocate grain across provinces, counties, and villages. As we have seen, this capability was almost certainly impaired. Moreover, it must also be remembered that approximately 10 percent of farm households lived on state farms, and while these farmers had access to land and the ability to grow other crops, they, too, were dependent on the PDS for some part of the year and thus almost certainly experienced shortfalls as well.
In short, as the famine crested, the evidence emphasizing the crucial role of entitlements and distribution becomes more evident. The FAO/WFP assess- ment from November 1997 is worth citing at some length: “There is also mounting evidence that much greater polarity in food consumption exists in the population than perceived hitherto. Reasons why this is occurring include transport difficulties, geographical differences, where some provinces are bet- ter equipped to deal with shortages than others, greater access amongst rural communities than urban and differential access to assets and foreign remit- tances and the corresponding ability to purchase food from emerging, though relatively insignificant, ‘private’ markets.” As we will argue in chapter 7, these comments are remarkably prescient. As early as 1997, the first signs were visible that the famine and food shortages were driven not only by the collapse of the PDS but also by the emergence of differential capacities to command resources to purchase needed grain on the market.
Mortality I: Who Died?
We have looked at the famine through the lens of how and to whom food was distributed in North Korea. Another cut on these questions is to consider who died and how. The growing work on famine has provided some general insights
The Distribution of Misery 73
into this question that can be summarized quite succinctly (Devereux 2000). First, deaths occur not simply from starvation but from disease, as a result of either increased individual vulnerability or the simultaneous breakdown of public health systems and the unavailability of medicine. Second, the most vulnerable within the household are infants, children, and the elderly.22
These observations are borne out by the refugee interviews we have (par- ticularly Robinson et al. 1999, 2001).23 UN estimates for 1990–95 put infant mortality at 24.4 per 1,000 live births. The first Hopkins study finds under- four mortality to be nearly four times as high for the 1995–97 period (88.9 per 1,000). The second Hopkins study finds infant mortality of 57.4 per 1,000 and under-five mortality of 30.3. Both studies also find predictably high and elevated levels of mortality among the elderly as well. The 1998 Good Friends study, which asks about causes of death, underlines the significance not only of starvation but also of disease. Only 33 percent of respondents cited “starva- tion” as the cause of death; fully 51 percent ascribed death solely to disease and another 10 percent to starvation and disease.
In sum, the evidence we have on the distribution of mortality by age group confirms both prior research and the concerns of the donor community to focus on the children and elderly. This concern was compounded by an impor- tant institutional issue that became a point of conflict with the donors. Many of the most vulnerable populations were in institutions—orphanages and hos- pitals—that were less well positioned to defend their entitlements than were PDS centers or work units. Indeed, we have convincing evidence from the NGO community in particular that the government even sought to conceal the existence of these institutions.