**4. Discussion**

#### *4.1. Influence of Money and Grain Allocations on Regional Food Prices*

Food prices directly reflect the food supply and demand, and are negatively correlated with the crop yields, and are also the indicators of social stability [50]. Therefore, variations in food prices in disaster areas and non-disaster areas can indicate the effects of regional interactions and coordinated responses to the extreme drought events. Wheat prices from 1876 to 1879 were selected from the Food Price Database in the Qing Dynasty [41] for the analysis, as wheat has always been the primary food staple in China. Figure 7 shows the lowest and highest wheat prices in different provinces and regions in each year from 1876 to 1879.

**Figure 7.** Provincial wheat prices in 1876–1879. Blue lines: variations of the highest annual wheat price in the province.

The wheat prices in every province of the disaster area had risen with fluctuations. In Shanxi and Henan, the prices in 1877 and 1878 could be over four or five times higher than usual. Wheat prices in Shanxi fluctuated most violently, and the extent of the rise was the most significant. In 1878 and 1879, wheat prices in Henan soared. In Shandong and Hebei, however, the wheat prices were relatively stable, with just small increases. A positive correlation can be observed between wheat price variations and the severity of droughts and famines.

The relief silver and grain alleviated some of the damages caused by the famine, but the effects were still insufficient. Taking Shanxi as an example, from 1876 to 1879, Shanxi had received 1.76 million piculs of grain and 13 million taels of silver in total. The relief silver could purchase about 2.6 million piculs of wheat according to the average annual wheat price (5 taels of silver per picul) in Shanxi at that time. That is to say, on average, about 1.09 million piculs of wheat were distributed to Shanxi each year. This amount of wheat could support 650,000 famine victims to survive for 8 months after the autumn harvest failure till the next summer harvest, based on the average daily food consumption of approximately 0.007 piculs per person [29]. However, at that time, the number of famine victims in Shanxi had exceeded 4 million [51], which means only 16% of the famished people could be supported. The situation remained severe, and some starving people even started resorting to cannibalism for survival. According to the records, starvation cannibalism occurred in respectively 42 worst-hit counties of Shanxi, 21 in Henan, and 11 Shaanxi [52].

In the non-disaster area, there were also obvious fluctuations in wheat prices. During 1877–1878, the harvest rates in Hubei, Hunan, and Jiangxi were just around 60–70% (Table 3). Sending relief grain to other provinces caused insufficient domestic supply, resulting in the sharp increases in wheat prices in Hunan and Jiangxi. On the one hand, inter-regional grain transfers worked effectively in stabilizing food prices in the disaster area. On the other hand, to some extent, the grain transfers disturbed the conditions of the food markets in the non-disaster area.

#### *4.2. Influence of Famine-Related Migrations on Regional Social Stability*

The regional interactions in social responses to the famine also profoundly influenced social stability. In 1877, more social unrest events (≥2–4/10,000 km2) took place in Henan, Shandong, Hebei, and Beijing (Figure 8), which were the main origins and destinations of the starving migrants. Very high unrest events density (≥6–7/10,000 km2) was found in Hebei. There were many reports of banditry and food robbery at the junction of Shaanxi, Shanxi, and Henan provinces. In 1878, the popular destination of famine migrants changed from the southern Jiangsu to Anhui, which made the social order of the former begin to improve. The places with a high density of social unrest events (≥2–4/10,000 km2) were recognized in Henan.

(**a**)

**Figure 8.** *Cont.*

 **Figure 8.** (**a**) Social unrest events density in 1877. (**b**) Social unrest events density in 1878.

When the famine victims left from the worst-hit areas to the surrounding slightly-impacted areas, or from villages to towns and cities, the social unrest events also "followed" with them from the disaster areas to the destinations for migrants. Thus, for the non-disaster areas, the managemen<sup>t</sup> of the starving migrants was a challenge of regional governance.

#### *4.3. Regional Interaction Responses and Transfer-Dispersion of the Impacts of Extreme Weather Events*

Famine-related migration represents the dispersion of the population pressure in the famine-struck areas (Figure 9). The spatial redistribution of the famine victims dispersed the population pressure in their places of origin, but increased the population pressure in their destinations. Famine-related migration is also connected to social unrest events and represents the dispersion of the social impacts of the extreme drought events.

The allocation of money and grain is an administrative action, with the purpose of adjusting the gaps in food production between disaster areas and non-disaster areas, and increase the opportunities to obtain food for the victims in the disaster areas. Affected by the allocation, food prices in some of the non-disaster areas also increased. To a certain extent, the allocation of money and grain played a role in transferring the social impacts of disasters and had positive effects on the post-disaster recovery (Figure 9).

In Chinese history, normally there were two areas of destination for the migration driven by the events related to climate change. One is the southeastern part of China, with warm and humid weather and fertile soils. The other is the arid and barren northwest. Migrating to the northwest seems to go against common sense, but it was because of the increase of conflicts between farmers and herders, and the increase of invasion of the northern nomads in the dry period [53]. Between 1876 and 1879, extreme droughts led to a server famine in North China, and many people migrated in search of food. However, in spite of the food production, the starving migrants selected their destinations using the location and proximity to their homeland as the determining factors. This is because that the famine-related migrations at that time were mainly temporary, and it was difficult for starving people

to complete long-distance travel. Compared with the climate-driven migration mentioned above, the migration during the North China Famine of 1876–1879 was a spontaneous social response to the extreme disaster events and the dispersion of short-term population pressure in the disaster areas.

**Figure 9.** Process of regional interaction responses and transfer-dispersion of the impacts of disasters.

Based on the vast canal network, the central governmen<sup>t</sup> planned, organized and carried out the allocation of money and grain, which was also the process of impact transfer across the whole country. In this network, the relationship between regions was more complex than that in the migration network. This time, in spite of the distances, the grain storage and accessibility of a place were the main factors in the decision-making process.
