A comparative study of collective relocation projects for poverty alleviation and earthquake reconstruction in rural China in terms of social integration

Figures & data

Table 1. General comparison between PAR (A) projects and ERR (B) projects.

Implementation area	PAR: nationwide in rural area, mainly in under-developed western and central China, especially in the key nation-level and province-level poverty-stricken counties. ERR: in severely damaged disaster-stricken areas, in which mountainous area is the most severely affected area due to significantly larger in the intensity of casualties and economic losses. Influenced by geology and terrain, the extensive hilly areas are not only disaster-stricken but also have a concentration of poverty (Gao, Chen, and Fan 2010).
Mode adoption	PAR: mainly decided by policy factors, comprehensive integrate and allocate resources to meet the region’s medium and long-term development goals. The distribution of resources per community is proportional to the severity of the devastation (Yu, et al., 2008). ERR: mainly based on condition factors, namely the geological condition of the original site after the disaster and the degree of damage to the houses2008.
Construction content	ERR has a stage of “restoration” before a stage of “rejuvenation”, including emergency shelter and prevention education, etc. (Sun et al. 2010; Hirayama 2000), while PAR doesn’t. ERR has relatively higher housing construction standards than PAR in terms of per capital area and investment (Tan 2019). ERR has more concerns in improving the ability to resist future disasters and enhancing communities’ safety and resilience (Alexander 2012).
Public engagement	ERR: after Lushan earthquake, three dominant approaches were identified as general principles, namely unified planning/unified reconstruction (UPUR), unified planning/joint reconstruction (UPJR), and unified planning/self-reconstruction (UPSR) (Yang, Wang, Huang, 2020). PAR: voluntary, fully planned, mainly unified planning/unified reconstruction (UPUR) headed by the government due to the poor households’ limited social capital, resulting in insufficient public engagement (Lo and Wang 2018). ERR has a generally higher level of public engagement and information access, it has a higher “emergency” level due to sudden displacement.
Challenge	PAR: re-settlers’ low relocation willingness, returning to original places, returning to poverty (Zhen 2019). ERR is confronted with an additional challenge of “mental reconstruction” (Akerkar and Fordham 2017).

Table 2. The four dimensions and the calculation method for the resettlement typology.

	Dimensions	Difinition	Description in Resettlement	Formula	Indication
X1:	Exposure	Potential contact between two groups	Population X (re-settler) proportion	PPx = X/T	Relocation intensity
X2:	Concentration	Physical space occupied by a group	X land use area ratio	LARx = Ax/A
X3:	Clustering	Contiguous space a group occupies	Average distance between residents	Ptt = $\sum _{i=1}^n\sum _{j=1}^n{t}_i{t}_j{d}_{ij}/T^2$	Spatial layout
X4:	Centralization	Relative proximity to the center	Relative proximity to village center	RCE = Px/Py-1	Relative location

Group X: re-settler group; Group Y: local group; X: total population of the re-settler group; Y: total population of the local group; T: total population;

$t_{i/j}$: total population in unit i/j; $d_{\mathrm{i}\mathrm{j}}$: the distance of centroids of units i and j

A: total construction land-use area of village; Ax: land-use area occupied by re-settlers; Ptt: average distance between all residents regardless of the group

RCE: Px and Py are X group and Y group’s average proximity to the village center, respectively.

$\mathrm{P}\mathrm{x}=\sum _{i=1}^n{x}_i\mathrm{e}\mathrm{x}\mathrm{p}(-d_{ij})/\mathrm{X}$; $\mathrm{P}\mathrm{y}=\sum _{i=1}^n{y}_i\mathrm{e}\mathrm{x}\mathrm{p}(-d_{ij})/\mathrm{Y}$, in which, $x_{i}and{y}_i$ are the population of X group and Y group in unit i.

Units were defined using natural element boundaries, land use, and the road network.

Figure 1. Distribution charts of the PAR and ERR cases in the four dimensions.

Figure 2. Maps of cases of the enclave type.

Figure 3. Maps of cases of the adjacent type.

Figure 4. Maps of cases of the centralized type.

Figure 5. Maps of cases of the infill type.

Figure 6. Definition and calculation method of socio-spatial indices.

Table 3. Basic information of the study cases.

Typology		Case	Move-in year	Sample size
				Total	Re-settler	Local	% (re-settler)
PAR	Adjacent	Shenjiayin	2017	107	43	64	40.2%
	Enclave	Xiling	2016	122	40	82	32.7%
	Centralized	Kongque	2017	97	50	47	51.5%
	Infill	Qinjiahe	2017	128	48	80	37.5%
Total				454	181	373	39.9%
ERR	Adjacent	Chaoyang	2016	98	33	65	33.6%
	Enclave	Yujia	2016	123	36	87	29.2%
	Centralized	Feixian	2016	126	65	61	51.6%
	Infill	Yanxikou	2016	91	34	57	37.5%
Total				438	168	270	38.3%

Table 4. Enclave type comparison.

SO	Group	Village		Mean SO value	Mean difference (A-B)	Std. Error	P
	Re-settler	PAR	Xiling	0.892	0.151***	0.027	0.000
		ERR	Yujia	0.741
	Local	PAR	Xiling	0.910	0.045**	0.018	0.016
		ERR	Yujia	0.865
SP	Group	Village		Mean SP value	Mean difference (A-B)	Std. Error	P
	Re-settler	PAR	Xiling	0.975	0.009**	0.003	0.009
		ERR	Yujia	0.966
	Local	PAR	Xiling	0.987	0.013***	0.002	0.000
		ERR	Yujia	0.974

*p < 0.05; **p < 0.01; ***p < 0.001

Table 5. Adjacent type comparison.

SO	Group	Village		Mean SO value	Mean difference (A-B)	Std. Error	P
	Re-settler	PAR	Shenjiaying	0.744	−0.038	0.027	0.158
		ERR	Chaoyang	0.782
	Local	PAR	Shenjiaying	0.827	−0.019	0.020	0.359
		ERR	Chaoyang	0.846
SP	Group	Village		Mean SP value	Mean difference (A-B)	Std. Error	P
	Re-settler	PAR	Shenjiaying	0.926	0.051**	0.016	0.001
		ERR	Chaoyang	0.875
	Local	PAR	Shenjiaying	0.924	−0.010	0.012	0.405
		ERR	Chaoyang	0.934

Table 6. Centralized type comparison.

SO	Group	Village		Mean SO value	Mean difference (A-B)	Std. Error	P
	Re-settler	PAR	Kongque	0.660	0.003	0.029	0.913
		ERR	Feixian	0.657
	Local	PAR	Kongque	0.700	−0.023	0.030	0.433
		ERR	Feixian	0.723
SP	Group	Village		Mean SP value	Mean difference (A-B)	Std. Error	P
	Re-settler	PAR	Kongque	0.929	−0.023**	0.010	0.031
		ERR	Feixian	0.952
	Local	PAR	Kongque	0.880	−0.010	0.011	0.405
		ERR	Feixian	0.870

Table 7. Infill type comparison.

SO	Group	Village		Mean SO value	Mean difference (A-B)	Std. Error	P
	Re-settler	PAR	Qinjaihe	0.760	−0.011	0.716	0.716
		ERR	Yanxikou	0.771
	Local	PAR	Qinjaihe	0.747	−0.040	0.095	0.095
		ERR	Yanxikou	0.787
SP	Group	Village		Mean SP value	Mean difference (A-B)	Std. Error	P
	Re-settler	PAR	Qinjaihe	0.885	−0.009	0.533	0.533
		ERR	Yanxikou	0.894
	Local	PAR	Qinjaihe	0.870	−0.037**	0.001	0.001
		ERR	Feixian	0.870

Table 8. Rankings of mean SO and SP value in the four types and two groups respectively in A(PAR) and B(ERR) projects.

Value	Group	Category	Rankings
			1	2	3	4
SO	Re-settler	PAR	Centralized	Adjacent ≈ Infill	Enclave	/
		ERR	Centralized	Enclave ≈ Infill ≈ Adjacent	/	/
	Local	PAR	Centralized ≈ Infill	Adjacent	Enclave	/
		ERR	Centralized	Infill	Adjacent ≈ Enclave	/
SP	Re-settler	PAR	Infill	Adjacent ≈ Centralized	Enclave	/
		ERR	Adjacent ≈ Infill	Centralized ≈ Enclave	/	/
	Local	PAR	Infill ≈ Centralized	Adjacent	Enclave	/
		ERR	Centralized	Infill	Adjacent	Enclave

If p > 0.05, we judged as statistically equal (“≈”)

Table 9. Summary of influencing social factors for the local group and re-settler group.

*The factors that only found significant for the PAR are in orange cells, the ones that only found significant for the ERR are in orange cells, the common factors are in white cells.

Table 10. Overall regression on social isolation degree (SO value) and spatial variables for all residents in A and B.

*p < 0.05; **p < 0.01; ***p < 0.001. The cells in black, dark grey, and light grey are representing the Top 3 most critical factors respectively.

Table 11. Regression on social isolation degree (SO value) and spatial variables for enclave type in A and B.

*p < 0.05; **p < 0.01; ***p < 0.001. The cells in black, dark grey, and light grey are representing the Top 3 most critical factors respectively.

Table 12. Regression on social isolation degree (SO value) and spatial variables for the adjacent type in A and B.

*p < 0.05; **p < 0.01; ***p < 0.001. The cells in black, dark grey, and light grey are representing the Top 3 most critical factors respectively.

Table 13. Regression on social isolation degree (SO value) and spatial variables for the centralized type in A and B.

*p < 0.05; **p < 0.01; ***p < 0.001. The cells in black, dark grey, and light grey are representing the Top 3 most critical factors respectively.

Table 14. Regression on social isolation degree (SO value) and spatial variables for the enclave type in A and B.

*p < 0.05; **p < 0.01; ***p < 0.001. The cells in black, dark grey, and light grey are representing the Top 3 most critical factors respectively.

Gao, X., T. Chen, and J. Fan. 2010. “Population Capacity in the Wenchuan Earthquake Reconstruction Areas.” Acta Geographica Sinica 65 (2): 164–176.

 Google Scholar

Kongjian, Y., X. Liang, L. Qing, L. Wenjin, and L. Dihua. 2008. “Post Quake Resettlement Scenarios: With a Case Study of Sichuan Wolong Special Administrative Region.” New Architecture 4: 76–79.

 Google Scholar

Sun, M. J., B. F. Chen, J. Z. Ren, and T. T. Chang. 2010. “Natural Disaster’s Impact Evaluation of Rural Households’ Vulnerability: The Case of Wenchuan Earthquake.” Agriculture and Argricultural Science Procedia 1: 52–61. doi:10.1016/j.aaspro.2010.09.007.

 Google Scholar

Hirayama, Y. 2000. “Collapse and Reconstruction: Housing Recovery Policy in Kobe after the Hanshin Great Earthquake.” Housing Studies 15 (1): 111–128. doi:10.1080/02673030082504.

 Web of Science ®Google Scholar

Tan, X. W. 2019. “The Evolvement of Poverty Alleviation for Resettled People in 70 Years: From 1949 to 2019.” Chinese Rural Economy 8 1–17 http://rdi.cass.cn/rdi_cssn_yjcg/201911/P020191115543950703436.pdf . (in Chinese).

 Google Scholar

Alexander, D. 2012. “What Can We Do about Earthquakes? Towards a Systematic Approach to Seismic Risk Mitigation.” 2012 NZSEE Conference, Christchurch, New Zealand.

 Google Scholar

Lo, K., and M. Y. Wang. 2018. “How Voluntary Is Poverty Alleviation Resettlement in China.” Habitat International 73: 34–42. doi:10.1016/j.habitatint.2018.01.002.

 Web of Science ®Google Scholar

Zhen, X. Y. 2019. Evaluation of the Effects of Targeted Poverty Alleviation Policies. China: National Institute of Development and Strategy, Renmin University of China. (in Chinese).

 Google Scholar

Akerkar, S., and M. Fordham. 2017. “Gender, Place and Mental Health Recovery in Disasters: Addressing Issues of Equality and Difference.” International Journal of Disaster Risk Reduction 23: 218–230. doi:10.1016/j.ijdrr.2017.03.014.

 Web of Science ®Google Scholar