Post-Disaster Recovery and Economic Impact of ... - Earthquake Spectra

Post-Disaster Recovery and Economic Impact of Catastrophes in China Jidong Wu,a), b) Ning Li,b) Wei Xie,a) Yang Zhou,a) Zhonghui Ji,a) and Peijun Shib)

This paper implements a temporal–spatial recovery measurement of the catastrophic 1976 Tangshan earthquake using available statistical data. The results show that the gross regional product (GRP) level of the Tangshan region achieved a new normality after seven years. During this recovery process, net indirect losses totaled RMB3.7 billion and net indirect gains totaled RMB3.9 billion at the 2007 price level. The area surrounding the Tangshan region benefited from the disaster, both in terms of GRP level and per capita GRP level, at least in the short term. The sector-level economic recovery process seems longer. The production level of the construction sector was 0.9 to 2.5 times that of the pre-disaster level during its 11-year recovery period. The per capita GRP level of the Tangshan region was 1.7 times that of pre-earthquake 30 years later. This quantitative disaster recovery analysis is critical for validating or initializing economic loss estimation models. [DOI: 10.1193/090511EQS221M]

INTRODUCTION The question of whether the economic effect of a disaster is negative or positive has been the focus of traditional economic analysis of disasters (Dacy and Kunreuther 1969, Cochrane 1975). In a short-run study of the effects of disasters, Albala-Bertrand (1993, p. 104) claimed that the indirect economic effects of disasters are “more a possibility than a reality.” A localized disaster may not have a negative net effect on the macro-economy if negative impacts from damage are canceled out by the positive impacts of recovery and reconstruction activities (Albala-Bertrand 2007). Interestingly, Noy (2009) finds a negative correlation between disasters and the long-run economic growth rate: natural disasters will typically cause a drop in output of nine percentage points in developing countries. A long-run empirical study by Skidmore and Toya (2002) showed that climatic disasters are positively associated with higher long-run economic growth, while geologic disasters are negatively correlated with growth. In the studies described above, the quantitative estimates of the size of the macroeconomic impact of disasters are questionable. First, the authors model natural disasters and countries as homogenous when the economic consequences induced by disasters will vary across the affected region (Ellson et al. 1984, Steenge and Bočkarjova 2007,

a)

Key Laboratory of Environmental Change and Natural Disaster, Ministry of Education of China, Beijing Normal University, Beijing 100875, China b) State Key Laboratory of Earth Surface Processes and Resource Ecology, Beijing Normal University, Beijing 100875, China, Correspondence to: Peijun Shi ([email protected]) 1825

Earthquake Spectra, Volume 30, No. 4, pages 1825–1846, November 2014; © 2014, Earthquake Engineering Research Institute

1826

WU ET AL.

Strobl 2012). The low quality and inconsistency of disaster data are also a concern. Moreover, most of the macroeconomic impact studies were based on cross-country data; while a positive or negative impact of disasters for developing or developed countries can be estimated, this is insufficient for regional disaster risk management because such estimations cannot provide differential predictions at the sectoral or interregional scale to local governors. Significant progress has been made in recent years in regional economic impact modeling of disasters, especially “Modeling Spatial and Economic Impacts of Disasters,” edited by Okuyama and Chang (2004). This book shows how to use economic models, such as input–output and computable general equilibrium models, to estimate the economic impact of disasters. Santos and Haimes (2004) proposed a modified input–output model able to reflect the extent of an economy’s inoperability post-disaster. Rose and Liao (2005) illustrated how a computable general equilibrium model can be used to estimate the sectoral and regional economic impacts of a lifeline supply disruption in the aftermath of a major earthquake. Tatano and Tsuchiya (2008) further developed a spatial computable general equilibrium model to estimate the economic loss due to seismic transportation network disruption. Hallegatte (2008) proposed the adaptive regional input–output (ARIO) model and demonstrated that indirect losses increase nonlinearly with direct losses. The ARIO model was also used to estimate the economic impacts of Hurricane Katrina, the storm surge risks in port cities (Hallegatte et al. 2011), and the Wenchuan earthquake (Wu et al. 2012). The results from input–output models and computable general equilibrium models can reflect the heterogeneous effects of disasters well. Such models have estimated that the economic impact on the disaster-affected areas is usually negative; the indirect economic losses were several times larger than the direct losses in a disaster scenario analysis (Hallegatte 2008). However, these economic impact estimation models usually lack sufficient validation due to a scarcity of observed economic statistics and disaster impact data. Moreover, the estimated economic impact was usually a net total impact; that is, it did not distinguish between the negative impacts from disaster damage and the positive impacts from recovery and reconstruction activities. Empirical measurement of positive and negative disaster impacts on different spatial scales is necessary and critical for improving the accuracy of the estimates from these economic models. Few studies have made such an effort, especially in the context of developing countries. The recovery timeframe is the focus of most studies on disaster loss, as the duration of disaster recovery is directly related to the value of indirect economic losses and the total economic costs caused by the disaster (Brookshire et al. 1997, Rose et al. 1997). Disaster recovery research has transformed the development history from one that uses qualitative descriptions (Hass et al. 1977) to one that involves quantitative statistics (Chang 2010). Hass et al. (1977) were the first to propose a conceptual framework for describing the disaster recovery process. They divided it into four stages: the emergency period, restoration phase, replacement-reconstruction phase, and developmental reconstruction period. This framework was used to predict the post-disaster recovery timeframe of Hurricane Katrina in New Orleans (Kates et al. 2006). Porter et al. (2001) attempted to estimate the repair time given various levels of building damage. Burton et al. (2011) tried to evaluate the spatial recovery of the rebuilt environment in Mississippi after Hurricane Katrina using repeated photography and found recovery disparities across different communities. Chang (2010)

POST-DISASTER RECOVERY AND ECONOMIC IMPACT OF CATASTROPHES IN CHINA

1827

proposed an exact quantitative measurement framework for urban disaster recovery using statistical data; this framework allows an estimation of the recovery duration of a given disaster. A comparison of “with- and without-” disaster values for various economic variables (e.g., per capita income) can also trace the effects of disasters in the recovery timeframe (Ellson et al. 1984). McComb et al. (2011) suggest that the adjacent regions surrounding the area affected by Hurricane Katrina may sustain lasting negative economic consequences. This persistence exists even though these surrounding areas can benefit from providing substitute production and shelter for the disaster-hit area in the short run. Ewing et al. (2005) find that the natural rate of unemployment in Corpus Christi, Texas, decreased after Hurricane Bret in 1999. Xiao (2011) examined the local economic impacts of the 1993 Midwest flood using time-series analysis and found that significant drops in personal income can be observed in the year of the event; however, the long-run effects seemed to be negligible. This article contributes to the literature on the economic effects of natural disasters by estimating the economic impacts of the disaster recovery process after the 1976 Tangshan earthquake in China using observed time-series data. We do so not only by measuring the economic recovery of the worst-hit area and the surrounding areas but also by distinguishing the negative impact (indirect losses) of disaster damage from the positive impacts (indirect gains) of recovery and reconstruction activities. To investigate the disaster recovery pattern in China, the recovery of the Tangshan region is also compared with that of Sichuan after the 2008 Wenchuan earthquake. The disaster impact mechanisms are also decomposed to provide guidance for disaster risk management. This case study can be considered a supplement to other disaster case studies focused on developing countries where there is no insurance. DATA AND METHODS STUDY AREA AND THE 1976 TANGSHAN EARTHQUAKE

The study area is the Tangshan region and the rest of the Hebei Province of China (Figure 1). The Tangshan region is one of the districts in Hebei Province. In 1976, the Tangshan region included Tangshan City (city area or downtown center) and ten counties: Fengnan, Fengrun, Luanxian, Luannan, Laoting, Tanghai, Qianan, Qianxi, Zunhua, and Yutian. Since 2002, Fengnan and Fengrun have also been placed under the jurisdiction of Tangshan City. On 28 July 1976, a moment magnitude 7.5 earthquake occurred below Tangshan City, which is now known as the site of the 20th century’s most deadly natural disaster. The death count from the Tangshan earthquake was 242,469. Of these deaths, 89.6% occurred in the Tangshan region (including 12,248 migrant deaths); an additional 167,539 people were severely wounded and 541,463 people suffered minor injuries in the region (Zou et al. 1997). The direct economic loss from the Tangshan earthquake was estimated to be around US$10 billion in 1976 (Grossi et al. 2006). No foreign aid for reconstruction was received at that time. Figure 2 shows the casualties and damage from field surveys performed by Zou et al. (1997) in the most severely damaged area of Tangshan City after this earthquake. The death toll in Tangshan City alone reached 135,919, accounting for 12.8% of its total population. Of these deaths, 54.8% were female, 34.6% were persons under age 15, and 16.2% of the deaths

1828

WU ET AL.

Figure 1. Study area. 60

Population (10 thousand)

Tangshan City

50

Minor injuries

Tangshan excl. Tangshan City

24.2%

Floating population in Tangshan

40 9.8%

Injures

30 20

Male

12.8%

Deaths

10

Female

0

0

Deaths

Injuries

5

10 15 20 Casualty rate (%)

Minor injuries

(a)

25

30

(b) Electricity

Age group

Major industrial buildings

Chemical industry

X T1 , are the net indirect gains in the recovery process. These are definitions as used in this paper. Depending on the geographic scale on which the economic assessment is performed, indirect losses and indirect gains may be different. For the disaster-affected region, damage in critical intermediate sectors may result in negative “ripple effects” in the economy. For example, the interruption of electricity, gas, and transportation has immediate consequences on the entire local economic system, even outside the immediate area of the disaster. The recovery measurement frameworks are used to identify the recovery process and recovery periods of the Tangshan region and the rest of Hebei (shown in Figure 1). Aggregate economic losses and gains are calculated to analyze the economic consequences of this quake.

1832

WU ET AL.

RESULTS The recovery period estimates for recovery variables related to the population and economy of the Tangshan region, based on the recovery measurement methods mentioned above, are given in Table 1. POPULATION RECOVERY

Figure 4a shows the long-term population trends of six seriously damaged counties suffering notable losses in population. These counties took one to eight years to return to their pre-disaster total population level. A disparity in the population recovery period is obvious. Fengnan County took eight years to regain its pre-earthquake population size due to large population losses. In contrast, Tangshan City (the worst-damaged area; 56.0% of the total death toll of the earthquake happened here) took only four years to regain its population size because of an influx of a large number of migrants from all walks of life nationwide1 (Mitchell 2008); reconstruction work also attracted a lot of labor force from rural area and other places (Wang and Sun 1996). Net immigrants reached almost 294,000 in Tangshan City from 1977 to 1986 (Zou et al. 1997)2; and some of them became permanent residents for nearly ten years’ physical reconstruction work (Zou et al. 1997). These were complemented also by a natural population increase among the earthquake survivors since 1977 (Figure 4c), thereby guaranteeing the rapid population regrowth of the Tangshan City. Only three years elapsed before the whole area of the Tangshan region regained its lost population, as shown in Table 1. Though the population recovery was relatively rapid, it was twenty-three years before the sex ratio regained its pre-earthquake level in the Tangshan region (Figure 4b). Sex imbalance was worse in Tangshan region compared with the rest of China after the earthquake, especially in Tangshan City (Figure 2b), which suffered an especially high proportion of female deaths (Zou et al. 1997, pp. 207). The sex ratio in Tangshan region regained its preearthquake level in 1998. Tangshan City took 21 years to regain its sex balance, which then developed similarly to that of China. The spatial disparity in population recovery is also very obvious in Tangshan region. Figure 4d shows 25-year population changes for the 11 counties of the Tangshan region (normalized to China). Except for Tangshan City, these counties experienced lower population changes than that of China. It seems that the disaster accelerated the agglomeration of the population in large cities; the annual natural population growth rate of Tangshan City was basically higher than that of the whole the Tangshan region just following the 1976 Tangshan earthquake (Figure 4c). Urbanization also contributed to this disparity, as Wu (1994) pointed that urban population experienced a rapid growth after the 1980s in China, and 80% of it was due to rural to urban migration. Moreover, Tangshan was one of the first few cities to 1

There were many external aid groups that participated in the reconstruction process from 1976 to 1986, including more than 100,000 Chinese People’s Liberation Army, more than 200,000 counterpart support workers, more than 20,000 medical aid and epidemic prevention people, and about 110,000 technical staff (Wang and Sun 1996, pp. 178). 2 The floating population is not counted in the net immigrants; this large-scale population change is rare in Chinese history (Zou et al. 1997).


Table 1.

List of recovery periods for different recovery variables in the Tangshan region Recovery variable

Population

Economy

1833

Definition

Recovery criteria Regaining its pre-disaster level in quantity

Recovery periods (year)

Population

Aggregate population

Sex ratio

Male per 100 female

GRP

GRP as a share of China’s GDP (%)

Per capita GRP

Per capita GRP as a ratio to China’s GDP per capita

7

GRP-industrial sector

Sector-output share as ratio to China

9

GRP-construction sector


11

GRP-tertiary sector


12

Achieving a new normalcy

3 23 7

implement urban housing reform in China (Shaw 1999), which allowed urban residents to build and rent their houses, and individual initiatives were brought back. It was also a government policy to promote housing reconstruction post-earthquake (Yu and Su 2003). Tangshan’s new residential construction rate in 1986 was 2.3 times that of the previous one in 1975 (Yu and Su 2003). The economic background and urbanization might be the mainly factors contributed to the long-term population trend. The government shifted their focus to securing an outstanding place for Tangshan in the national economic system after physical reconstruction was completed in 1986 (Mitchell 2008): The local living environment (including industrial pollution, building quality, and the landscape of the city) was obviously improved (Wang and Sun 1996). The level of urbanization began to rise steadily since the 1978 reforms in China (Ma 2002), which was also a background for the accumulation of urban population in Tangshan city. ECONOMIC RECOVERY

GDP, together with its regional equivalent GRP, can be used as an indicator for measuring the output change of economic activities. Figure 5 shows the annual GRP and the per capita GRP change for the Tangshan region and the rest of Hebei from 1952 to 2008. To filter out exogenous influences unrelated to the disaster (e.g., price inflation), economic indicators were described as a ratio of the value of the Tangshan region to that of China. This measure should reflect the real effects of the earthquake on the local economic system, assuming that the impact of the Tangshan earthquake to China overall was minimal; we discuss the validity

1834

WU ET AL.

(a)

(c)

(b)

(d)

Figure 4. Temporal-spatial population recovery of the Tangshan region. (a) Population trends for six of eleven jurisdictional counties of the Tangshan region. (b) Sex ratio change comparison between the Tangshan region and China. (c) Natural population growth rate in percent. (d) Spatial disparity in population recovery (Population data shown in the map were calculated as the population change ratio of a county from 1975 to 1999 divided by the ratio of the demographic changes of the same period in China). (Data source: Tangshan Municipal Bureau of Statistics (TBS 1992, 2008, 2009), Hebei Provincial Bureau of Statistics (HBS 2009), and National Bureau of Statistics of China (NBS 2010).)

of this assumption later in the text. Economic recovery periods for different variables are given in the fifth column of Table 1. For the Tangshan region, the GRP level (defined as GRP as a share of China’s GDP) achieved a new normality in 1982, after having decreased by 14.3% (dropped from being 0.7% of the China’s total GDP in 1976 to 0.6% in 1975) in the year that the earthquake occurred3. After two years of reconstruction (or restoration), the GRP level first surpassed its pre-earthquake level (increased to 0.8% of the China’s total GDP in 1978, compared with 0.7% in 1975) then rapidly grew over the following five years (it was 105.4% of the preearthquake level on average). Over the next eight years (until 1990), there were fluctuating declines. However, a gradual growth trajectory began in 1991 and was maintained thereafter. The economy appears to have redeveloped since then. In 2005, the Tangshan region’s GRP level was 1.5 times larger than it was in 1975 prior to the earthquake. The growth ratio of per capita GRP experienced a trajectory similar to, but steeper than, that of total GRP: by 2005, 3

The production level change of Tangshan was based on the baseline level pre-disaster in 1975, the same below for the GRP per capita level change.


GRP-Hebei GRP-Rest of Hebei GRP-Rest of Hebei excl. construction

1835

GRP per capita-Tangshan region GRP per capita-Hebei GRP-Tangshan region GRP-Tangshan region excl. construction

6 2.4

5 4.5

2

4 3.5

1.6

3 2.5

1.2

2 1.5

0.8

GRP per capita as ratio to China

GRP as percent of China (%)

5.5

1 0.5 0.4 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004 2008

Figure 5. Economic recovery of the Tangshan region. (Data source: Tangshan Municipal Bureau of Statistics (TBS 1992, 2008, 2009), Hebei Provincial Bureau of Statistics (HBS 2009), and National Bureau of Statistics of China (NBS 2010).)

the per capita GRP level (expressed as a ratio to China’s GDP per capita) was 1.7 times larger than it was prior to the earthquake in 1975. The economic reform in China beginning from 1978 and taking a strong hold in the mid1980s was an important aspect that promoted the economic recovery process. Since 1978, China has undergone a series of radical economic reforms that lead to a transition from central planning to a mixed economy (Shirk 1993). China’s economic reforms started with improving micromanagement institution from 1978 to 1984, moved to reforming the rigid centrally planned resource-allocation system from 1984 to 1991, and, finally changed the overall macro-policy environment and developed a market economy from 1992 onwards, these reforms have improved economic incentives and efficiency, lead to adjustment of economic structure and rapid economic growth (Lin et al. 2003). Tangshan, as a traditional industrial city, with its proximity to the coastal line and skilled labor in large quantities, which added further heft the economic boom evidenced by the GRP level in Figure 5, in turn helped to accelerate the reconstruction process (Mitchell 2008), is among the regions that benefited most by the reform (Yao 2009). The State Council approved Tangshan city as one of the 13 national “comparatively big” cities in 19844. The Tangshan region was further 4

The State Council of the people’s Republic of China approved 13 cities (including Tangshan, Dalian, Datong, Anshan and Chongqing) as national “comparatively big” cities by the end of 1984. These cities can draft local laws and regulations as they needed (http://www.people.com.cn/item/flfgk/gwyfg/1984/112103198403.html, retrieved on 7 March 2013).

1836

WU ET AL.

listed in the Beijing-Tianjin-Tangshan economic zone in 1986, and then crossover into the 24 cities that the GRP surpassed RMB10 billion in 1988, and became one of the open coastal regions approved by the State Council in the same year (TBS 2008). The effective use of exports and foreign investment were the main contribution to the superb growth rates of coastal areas relative to the national average (Wei 1994, Liu et al. 2002). Furthermore, disaster damage and related reconstruction provide opportunities to update the capital stock and adopt new technologies (Skidmore and Toya 2002, Cuaresma et al. 2008); this “positive shock” was explained by the concept of Schumpeterian “creative destruction,” embedded in a theory of endogenous growth (Aghion and Howitt 1998). Figure 5 also shows that the stimulus to the construction sector from reconstruction is very obvious in the seven-year recovery period: the ratio of the Tangshan region’s GRP to Chinese GDP excluding the construction sector is clearly lower than the overall GRP-GDP ratio during this period. This expansion, mainly owing to reconstruction demand especially in the construction sector, is common after disasters. A prominent example of this phenomenon is the Kobe earthquake (Chang 2010). If we accumulate the GRP losses and gains over the seven-year recovery process using the 2007 price level, the net indirect losses due to the Tangshan earthquake reached approximately RMB3.7 billion, and the net indirect gains were RMB3.9 billion. Compared with the economic recovery process of the Tangshan region from 1976 to 1982. The GRP levels of the rest of Hebei province and Hebei as a whole were above their pre-earthquake levels (Figure 5). If we accumulate the indirect economic effects using the 2007 price level, the net indirect gains of the rest of Hebei province reached RMB48.7 billion. The ratio of the GRP per capita of Hebei to the overall Chinese GDP per capita was also always higher than just before the earthquake in this period. Structural economic change is an important factor impacting the long-term economic development of the disaster-affected area. The economy of the Tangshan region developed from the primary sector to secondary and tertiary sectors based economies from 1968 to 2007 (Figure 6b). Figure 6a shows the sector-output share value variations in the Tangshan region from 1968 to 2007, which are also normalized to China overall. Here, we focus on the secondary sector of the economy, i.e., the industrial and construction sectors. The sector-output share value of the industrial sector did not significantly decrease while facing the challenges of such a large earthquake5. This outcome was attributed to the rapid restoration of the Tangshan region’s industrial production. As the “cradle of China’s modern industrialization” through the rejuvenation of the old industrial facilities (steel, railway rolling stock, chemicals, ceramic, and cement), the Tangshan region’s industrial production recovery was a high priority and attracted a large amount of resources from the central government during the physical reconstruction stage, and many industrial plants were established locally to produce reconstruction materials during this period (Chen et al. 1988). The stimulus to the construction sector from the reconstruction process can also clearly be observed in the sector output proportions from 1976 to 1982 (Figure 6a). Compared with the 5

Major industrial buildings and civil buildings were damaged in Tangshan City, the core of the Tangshan region, described above in Figure 2d, while more than half of the industrial added value of Tangshan region comes from Tangshan City according to the statistic data from Tangshan Municipal Bureau of Statistics Fen jin de Tangshan si shi san nian 1949–1992 (TBS 1992).


1837

2.5 2 1.5 1 0.5

Percentage of GRP (%)

Composition of GRP as ratio to China

60

Primary Sector Industry Construction Tertiary Sector

3

50 40 30 20 10

0 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004 2008 Year

0 19681972 19761980 19841988 19921996 20002004 2008 Year

(a)

(b)

Figure 6. Economic structure change pre- and post-earthquake in the Tangshan region. (a) Change of the composition of GRP as ratio to China. (b) Change of the percentage of GRP (the legend is the same as Figure 6a). (Data source: Tangshan Municipal Bureau of Statistics (TBS 1992, 2008, 2009), and National Bureau of Statistics of China (NBS 2010).)

seven-year recovery of GRP, the economic recovery across industrial sectors of the Tangshan region took even longer. Stabilization was attained in the industrial sector after nine years. The process was even slower in the construction (11 years) and the tertiary sectors (12 years). The production level in the construction sector was greater than its pre-earthquake level in 9 of the 11 years of the recovery process. During this period, the annual production level ranged from 0.9 to 2.5 times that of the pre-disaster period. The production level of the tertiary sector (103.7% of pre-disaster) was higher than the pre-disaster level, but production remained below pre-disaster levels for the industrial sector (93.8% of pre-disaster) in its nineyear recovery process, which first re-attained its pre-disaster level in 2001. The tertiary sector has maintained a development level above the pre-earthquake level since 1985, while the construction sector has developed at a level lower than the pre-earthquake level since 1995. A structural industrial transformation (or re-optimization) from construction to the other sectors occurred, for example, the contradictory trends between the primary and industrial sectors, and also the tertiary sector in the long run. The recovery of the Tangshan region was considered a surprise owing to a successful outcome that the recovery planners may not have initially anticipated: the rapid physical reconstruction and accelerated economic growth rates in the region (Mitchell 2008). The city received the Habitat Scroll of Honor Award from the United Nations in 1990 for its outstanding reconstruction of its communities (Yu and Su 2003). The Tangshan region has transitioned to a new, export-oriented economy (TBS 2008), new shallow and deepwater port facilities were built on the nearby Gulf of Bohai, and the new Tangshan region ranked high on the list of China’s top 50 most vibrant urban economies in 1996 (Yu and Su 2003). The strong centralized government leadership was a major factor in the successful recovery of Tangshan region that eventually surpassed pre-disaster economic conditions (Ge et al. 2010). This leadership gave priority to the disaster-stricken communities to receive public spending and tax relief, as well as the rapid raising of national resources and manpower for assistance (Mitchell 2008). The same factor played a significant role in the response to the 2008 Wenchuan earthquake (Wu et al. 2012). Aside from the importance attached to economic recovery in the aftermath of the earthquake, the urbanization and the socioeconomic transformation that has swept China since the 1980s also contributed

1838

WU ET AL.

to the Tangshan region’s successful recovery as described above. Entering the 21st century, Tangshan, as the heart of circum-Bohai Sea economic circle, possesses a combined advantage of port, resource, location, and industry. This coastal open city is maintaining its rapid economic growth; the average annual GRP growth rate reached 12.2% from 1979 to 2007 (TBS 2008). The experience of the Tangshan region’s economic recovery provides an initial estimate of the parameters for a disaster loss model for China and provides additional knowledge that can aid recovery planners in improving and enhancing response to inevitable future disasters. DISCUSSION COMPARISON WITH THE 2008 WENCHUAN EARTHQUAKE

The economic recovery pattern following catastrophe in China is worth exploring further; hence, we present some comparative analysis between the 1976 Tangshan earthquake and the 2008 Wenchuan earthquake. Figure 7 shows the economic recovery of the worst-hit region (including ten seriously damaged counties in Sichuan) of Sichuan province after the M s 8.0 Wenchuan earthquake, which occurred on 12 May 2008. This earthquake killed 69,226 people, left 17,923 missing, and caused a direct economic loss of RMB845 billion in 2008. Of these losses, 97.2% of the total dead and missing people, 42.9% of the collapsed houses, and 39.5% of the direct economic losses were from the worst-hit region (NCDR and MOST 2008). Unlike the region most affected by the Tangshan earthquake, the worst-hit region of the Wenchuan earthquake is mainly located in the mountainous area. The GRP level (as a percentage of Chinese GDP) of the worst-hit area of Sichuan decreased by 35.4% in 2008 compared to the 2007 level. After three years of reconstruction, the region had still not returned to its pre-earthquake GRP level, but the GRP level of the rest

0.7

3.5

0.68

3

0.66

2.5

0.64

2

0.62 GRP-The worst-hit region hit GRP-The hardest hit region region GRP-The hardest

1.5

GRP-Else of of Sichuan Sichuan GRP-Rest GRP-Sichuan GRP-Sichuan GRP per per capita capita of of Sichuan Sichuan GRP

1 0.5 0 1998

(a)

0.72

4

2000

2002

2004 Year

2006

2008

0.6 0.58 0.56

GRP per c apita as ratio to China

GRP as percent of China (%)

4.5

0.54 2010

(b)

Figure 7. Economic recovery of Sichuan after the 2008 Wenchuan earthquake. (a) The worst-hit region post-earthquake. (b) Economic recovery of Sichuan. (Data source: Sichuan Provincial Bureau of Statistics Sichuan Statistical Yearbook (Various years, 1999–2011) and the National Bureau of Statistics of China Statistical Yearbook (Various years, 1999–2011), published by China Statistics Press.)


1839

of Sichuan experienced a boom in those three years because of the reconstruction-demand stimulus. The GRP per capita level of Sichuan (as a ratio to Chinese GDP per capita) in 2010 increased by 10.0% compared with the 2007 pre-earthquake level. The economic recovery trajectory of the worst-hit region and surrounding areas (rest of Sichuan) of the Wenchuan earthquake is similar to the areas surrounding the Tangshan earthquake within the first three post-disaster years. However, the GRP of the worst-hit region of Sichuan needed more than three years to surpass its pre-earthquake level. The apparent change in the per capita GRP, compared to other economic variables, in both the Tangshan region and Sichuan just after the earthquakes deserves further discussion and explanation. First, the external recovery assistance post-disaster ensured a rapid recovery in the GRP of the disaster-hit area in the short term. Additionally, the population bonus for survivors, such as the increased per capita land resources resulting from the population decrease post-disaster, was an important factor that contributed to the improved per capita GRP in the Tangshan region in the short term. Simultaneously, as happened after the Wenchuan earthquake, labor shortages can occur after earthquakes, and the wage rise resulting from a labor supply shortage may also contribute an increase in per capita GRP. Furthermore, the growth rate of the local population during the earthquake recovery process in the Tangshan region was lower than the increase in GRP, leading to an apparent long-term improvement in the per capita GRP. Further studies are needed to analyze the per capita GRP change post-disaster over the long term. In understanding the difference in recovery curves or economic consequences, the spatial scale of the analysis provides the main explanation. By restricting the analysis to the worst-hit area (e.g., the worst-hit region in Sichuan for the 2008 Wenchuan earthquake natural) disasters can induce losses and gains in the local economy. On the one hand, the direct damage from the disaster can reduce production capacity, while on the other hand, reconstruction demand also increases post-disaster; the imbalance between supply shortage and demand expansion will induce bottleneck effects,6 and such economic interruption will cause economic losses. Even limited physical damage may have an enormous amount of economic disruption because of bottleneck effects within in the economic system (also known as interindustry ripple effects; Hallegatte 2008). For producers, in order to reduce the loss of profit, some undamaged firms may resolve to export their products to outside regions, while some firms may have to import goods from outside the region to meet their demand; both of these behaviors will increase transportation costs. For households, the price increase of consumer goods due to the supply shortage will increase their spending, and in turn will influence the consumption market. The severity of bottlenecks can vary over the recovery period, and the time needed for recovery plays an important role in determining the overall indirect losses. On the other hand, with outside assistance, reconstruction stimulus and preferential policies implemented by the central government can have positive effects (gains) on the local economy. The worst-hit region can also have a chance to re-optimize its economic structure 6

For the 1976 Tangshan earthquake, this industrial linkage impact in Tangshan maybe relatively lower than that of the Wenchuan earthquake in the worst-hit region of Sichuan, given that China in 1976 was not a market-based economy, and the economic structure was simple also (i.e., mainly concentrated in the primary sector and industry sector as Figure 6b shows).

1840

WU ET AL.

in the recovery process and potentially embark upon a sustainable growth trajectory. For the 1976 Tangshan earthquake, as Figure 6b shows, there is an explicit transformation among the inter-sectoral relationships among industries, especially in the shift from a low-income agricultural economy to a high-income urban-industrial economy. As the complexity of the economy becomes greater (Hewings et al. 1988), the relationship between sector linkages and economic development will become intricate too (Thakur and Alvayay 2012). However, when the cost of reconstruction spending is shifted to the region’s victims (e.g., when the economic conditions are back to “normal” with no preferential policies and outside aid), the gains from rebuilding are offset by reduced household spending (Cochrane 1997). For the surrounding areas (e.g., the rest of Hebei province for the 1976 Tangshan earthquake), if the worst-hit region produces the critical materials and substitutes, which are not available, then the ripple effects from the worst-hit area can spill over to the surrounding area; for example, undamaged firms in the surrounding region will have to reduce production due to a supply shortage. However, these negative effects seemed limited in the rest of Hebei in the short run. On the contrary, the rest of Hebei province obtained RMB48.7 billion in the seven-year economic recovery of the Tangshan region, as described above. For the Wenchuan earthquake, the GRP level of the rest of Sichuan from 2008 to 2010 was also higher than that of the pre-earthquake period. It seems that the reconstruction of the disaster-hit region (i.e., the Tangshan region) benefitted the economies of the surrounding areas (i.e., the rest of Hebei) in the short term. First, the supply shortage and reconstruction demand from the worst-hit region can be satisfied by speeding up production (increase work hours or production equipment) in the surrounding areas. Except for government guidance, this increased production is partly driven by profits because some commodity prices often increase due to the excess demand caused by disaster damage, such as the rise in building materials prices after the Wenchuan earthquake (Wu et al. 2012). Moreover, economic incentive policies implemented post-disaster usually benefited not only the worst-hit area, but also a less-damaged general region; for example, refer to the preferential tax policy implemented after the Wenchuan earthquake in the whole disaster region (Ge and Deng 2010). The spillover effect of these policies may be more useful for attracting investment and further stimulate its economy. Above all, indirect losses and indirect gains can be experienced by both the worst-hit area and by the surrounding area; the relative magnitudes depend on several constraints, including presence of outside assistance, preexisting economic development circumstances, the insurance and disaster relief system, and even previous disaster experience. For the Tangshan earthquake, the socioeconomic boom occurring in China at the time could have been a major factor in the successful economic recovery of the Tangshan region. For both the Tangshan earthquake and the Wenchuan earthquake, there was no earthquake insurance system; insurance payments for the Wenchuan earthquake were less than 0.3% of the direct economic losses (Wu et al. 2012). In the absence of insurance, the reconstruction funds mainly came from the government. After the Tangshan earthquake, the reconstruction investment from all levels of government (the central government, provincial government, and local government) reached RMB4.9 billion by the end of 1987 (Zou et al. 1997). After the Wenchuan earthquake, the government also implemented positive


1841

reconstruction policies to the disaster-hit area. Along with the reconstruction funds provided by the government, 18 assistance provinces (or cities) offered assistance with no less than 1% of their last ordinary budget revenues to their 18 counterpart counties (or districts) in Sichuan7. Thus, the government aid to the disaster areas ensured rapid economic recovery. Damaging effects can benefit the surrounding region, at least in the short run, as shown in the cases of the Tangshan earthquake and the Wenchuan earthquake. However, for the worsthit area, it is difficult to tell whether the disaster is beneficial or harmful to the local economy. While Guimaraes et al. (1993) propose that disaster areas will “potentially benefit from recovery efforts if there is a transfer of funds from outside the area that more than compensate for the losses,” further investigation is necessary to determine how these constraints drive the economic recovery across different spatial and temporal scales. LIMITATIONS AND FUTURE RESEARCH

This study offers some important insights into catastrophic disaster recovery and regional economic impacts of natural disasters. However, it has some limitations, and further disaster recovery case studies are necessary to better understand how natural disasters impact the economy at a regional scale. First, the net indirect losses and net indirect gains of the Tangshan region were calculated based on the controversial assumption that the counterfactual economic development level (i.e., the Tangshan region’s GRP as a share of China’s GDP) in the absence of the disaster is the same as that of the year just pre-disaster (i.e., in 1975 for the Tangshan earthquake). This crude assumption may exaggerate or reduce the GRP level without disaster, but it provides a base scenario for calculating the economic impact of disasters, and this assumption had also been used in estimating the economic impact of the 2005 Hurricane Katrina (Hallegatte 2008) and the 2008 Wenchuan earthquake (Wu et al. 2012). An econometric time-series model may be another alternative for predicting the economic development level without disaster. Xiao (2011) employed the autoregressive integrated moving average models to measure flood effects, but this estimate is strictly constrained to pre-disaster economic trends and linkages. The other required assumption is that a local disaster shock may not affect the macroeconomy negatively in any significant way for developing countries, which is supported by the finding of Albala-Bertrand (1993). This may not reflect the true economic consequences of the earthquake damage, especially just after the disaster, but it is useful to provide some indication of the magnitudes of the disaster effects on the regional scale. Second, the disaster data of the Tangshan region are taken from the report “Social Recovery and Social Issues after the Tangshan Earthquake” (Zou et al. 1997),8 which was a 7

In September 2008, four months after the earthquake, National Development and Reform Committee of China completed “The state overall planning for the post-Wenchuan earthquake restoration and reconstruction” to facilitate the reconstruction (Ge et al. 2010). 8 This work was done between 1990 and 1996, sociological approach was used for investigate the social recovery process of Tangshan region, and 10,060 valid questionnaires, related to thousands of recovery variables, were collected between 1990 and 1993 to study the social recovery issues (Zou et al. 1997, pp. 6). The work was led by Qijia Zou (from China Earthquake Administration), Ziping Wang (a disaster sociologists from Hebei United University), Feibi Chen (from China Earthquake Administration), and Shaoyu Wang (from Tangshan Engineering and Technology College). Twenty-seven scientists participated in the writing of this report.

1842

WU ET AL.

preliminary assessment of the social recovery from the Tangshan earthquake. Although some estimates have placed the death toll far higher than the government estimates—ranging from 250,000 to as high as 750,000 people (Hough and Bilham 2006)—statistically meaningful results can be obtained from these disaster data. More accurate disaster data from more recent disaster events (e.g., data from the Wenchuan earthquake) may be useful to further identify the post-disaster recovery process in China. Third, the population data from the statistical report do not include unregistered workers. The crudeness of this measure may influence its absolute value, but we are able to obtain a relatively reasonable population trend. Additionally, the economic data used in this study are all sourced from the statistical books. The reliability of Chinese statistical economic data was often questioned (Rawski 2001). Indeed, the statistical methodology change in China’s statistical system reform may contribute to this question, and results based on the statistical economic data should be treated with care. But although Chow (2006) pointed out that “Chinese official statistics are by and large reliable,” it does not mean that all analyses based on these data should be rejected, from an empirical perspective; the estimated effects are neither unreasonable nor inconsequential. Certainly, the recovery paths of differentially affected sectors widely vary. There is a need to know more not only about how post-disaster recovery proceeds over a particular timeframe, but also about how it varies across space. More detailed analyses of economic sectors or interregional scale studies may present interesting insights into disaster recovery research. One such example is the interregional input–output analysis of the electricity disruption simulation by Yamano et al. (2007) and the international dispersion of the economic effects of the 2004 Indian Ocean earthquake and tsunami studied by Okuyama (2010). CONCLUSIONS The present study provides an empirical analysis of the recovery of the catastrophic Tangshan earthquake at the urban and regional scales. It is specifically aimed at measuring the impact of catastrophic events inside and immediately outside of the disaster-hit region. The results showed that the Tangshan region took 3 years to regain its population size, but it took 23 years to regain its sex ratio. Meanwhile, the spatial population recovery disparity at the county level was very obvious even 25 years later, reflected in the movement of people from rural areas to Tangshan City. The GRP level (measured as GRP as a share of China’s GDP) of the Tangshan region achieved a new normality in seven years. Thirty years after the disaster, this measure was 1.5 times larger compared to the pre-earthquake level. At the same time, changes in industrial structure were also evident following the event, especially the movement from the primary to the tertiary sectors. A significant boost to the output level in the construction sector in the recovery process epitomized the larger economic recovery process; the production level of the construction sector was 0.9 to 2.5 times that of the pre-disaster level within the 11-year recovery period. The stimulus of reconstruction activities to the construction sector substantially contributed to the broader economic recovery of the Tangshan region. This earthquake caused RMB3.7 billion of net indirect economic losses but also brought RMB3.9 billion of net indirect gains within the seven-year recovery process. The area surrounding the worst-hit region


1843

(i.e., the rest of Hebei) benefited from the disaster, both in terms of GRP level and per capita GRP level. This empirical disaster recovery study can provide valuable knowledge for indirect economic loss assessment models or to design disaster recovery simulation models. It can also provide valuable insights and experiences for disaster managers preparing for the next unexpected disasters; for example, they can have an expectation about how long the recovery process will be if a similar damage extent is encountered in a future disaster. ACKNOWLEDGMENTS This work was supported by National Natural Science Foundation of China (No. 41101506 and No. 41171401), National Basic Research Program of China (No. 2012CB955402), and by International Cooperation Project funded by Ministry of Science and Technology of China (No. 2012DFG20710). Cordial thanks should be given to the three anonymous referees for their constructive suggestions and comments that greatly helped to improve the quality of this article.

REFERENCES Aghion, P., and Howitt, P., 1998. Endogenous Growth Theory, MIT Press, Cambridge, MA. Albala-Bertrand, J. M., 1993. The Political Economy of Large Natural Disasters: With Special Reference to Developing Countries, Clarendon Press, Oxford. Albala-Bertrand, J. M., 2007. Globalization and localization: An economic approach, in Handbook of Disaster Research (H. Rodriguez, E. L. Quarantelli, and R. R. Dynes, Eds.), Springer, New York, 147–167. Brookshire, D., Chang, S. E., Cochrane, H., Olson, R., Rose, A., and Steenson, J., 1997. Direct and indirect economic losses from earthquake damage, Earthquake Spectra 13, 683–701. Burton, C., Mitchell, J. T., and Cutter, S. L., 2011. Evaluating post-Katrina recovery in Mississippi using repeat photography, Disasters 35, 488–509. Chang, S. E., 2010. Urban disaster recovery: a measurement framework and its application to the 1995 Kobe earthquake, Disasters 34, 303–327. Chen, Y., Tsoi, K. L., Chen, F., Gao, Z., Zou, Q., and Chen, Z., 1988. The Great Tangshan Earthquake of 1976: An Anatomy of Disaster, Pergamon Press, Oxford, UK. Chow, G., 2006. Are Chinese official statistics reliable?, CESifo Economic Studies 52, 396–414. Cochrane, H. C., 1975. Natural Hazards and their Distributive Effects, Natural Hazards Research Applications Information Center, Boulder, CO. Cochrane, H. C., 1997. Economic Impacts of a Midwestern Earthquake, NCEER Bulletin, available at http://mceer-nt4.mceer.buffalo.edu/publications/bulletin/97/11-01/jan97.pdf, 11, 1–5. Cuaresma, J. C., Hlouskova, J., and Obersteiner, M., 2008. Natural disasters as creative destruction? evidence from developing countries, Economic Inquiry 46, 214–226. Dacy, D. C., and Kunreuther, H. C., 1969. The Economics of Natural Disasters, Free Press, New York.

1844

WU ET AL.

Ellson, R. W., Milliman, J. W., and Blaine Roberts, R., 1984. Measuring the regional economic effects of earthquakes and earthquake predictions, Journal of Regional Science 24, 559–579. Ewing, B. T., Kruse, J. B., and Thompson, M. A., 2005. Empirical examination of the Corpus Christi unemployment rate and Hurricane Bret, Natural Hazards Review 6, 191–196. Ge, Y., Gu, Y., and Deng, W., 2010. Evaluating China’s National Post-Disaster Plans: The 2008 Wenchuan Earthquake’s Recovery and Reconstruction Planning, International Journal of Disaster Risk Science 1, 17–27. Grossi, P., del Re, D., and Wang, Z., 2006. The 1976 Great Tangshan Earthquake: 30-year retrospective, Risk Management Solutions, Inc., Newark, CA, 1–14. Guimaraes, P., Helfner, F. L., and Woodward, D. P., 1993. Wealth and income effects of natural disasters: an economic analysis of Hurricane Hugo, The Review of Regional Studies 23, 97–114. Haas, J. E., Kates, R. W., and Bowden, M. J. (Eds.), 1977. Reconstruction Following Disaster, MIT Press, Cambridge, MA. Hallegatte, S., 2008. An adaptive regional input–output model and its application to the assessment of the economic cost of Katrina, Risk Analysis 28, 779–799. Hallegatte, S., Ranger, N., Mestre, O., Dumas, P., Corfee-Morlot, J., Herweijer, C., and Muir-Wood, R., 2011. Assessing climate change impacts, sea level rise and storm surge risk in port cities: a case study on Copenhagen, Climatic Change 104, 113–137. Hebei Provincial Bureau of Statistics (HBS), 2009. New Hebei: Sixty Years 1949–2009, China Statistics Press, Beijing. Hewings, G. J. D., Sonis, M., and Jensen, R. C., 1988. Fields of influence of technological change in input–output models, Papers of the Regional Science Association 64, 25–36. Hough, E., and Bilham, R. G., 2006. After the Earth Quakes: Elastic Rebound on an Urban Planet, Oxford University Press, New York. Kates, R. W., Colten, C. E., Laska, S., and Leatherman, S. P., 2006. Reconstruction of New Orleans after Hurricane Katrina: a research perspective, Proceedings of the National Academy of Sciences 103, 14653–14660. Lin, J. Y., Cai, F., and Li, Z., 2003. The China Miracle: Development Strategy and Economic Reform, Chinese University Press, Hong Kong SAR, China. Liu, X., Burridge, P., and Sinclair, P. J. N., 2002. Relationships between economic growth, foreign direct investment and trade: Evidence from China, Applied Economics 34, 1433–1440. Ma, L. J. C., 2002. Urban transformation in China, 1949–2000: a review and research agenda, Environment and Planning A 34, 1545–1569. McComb, R., Moh, Y.-K., and Schiller, A. R., 2011. Measuring long-run economic effects of natural hazard, Natural Hazards 58, 559–566. Mitchell, J. K., 2008. Including the capacity for coping with surprises in post-disaster recovery Policies: Reflections on the experience of Tangshan, China, Behemoth: A Journal on Civilisation 3, 21–38. National Bureau of Statistics (NBS), 2010. China Compendium of Statistics 1949-2008, China Statistics Press, Beijing. National Commission for Disaster Reduction (NCDR), and Ministry of Science and Technology of China (MOST), 2008. Wenchuan earthquake disaster—a comprehensive analysis and evaluation, Science Press, Beijing.


1845

Noy, I., 2009. The macroeconomic consequences of disasters, Journal of Development Economics 88, 221–231. Okuyama, Y., 2010. Globalization and Localization of Disaster Impacts: An Empirical Examination, CESifo Forum 11, 56–66. Okuyama, Y., and Chang, S. E., 2004. Modeling Spatial and Economic Impacts of Disasters, Springer, New York. Porter, K. A., Kiremidjian, A. S., and LeGrue, J. S., 2001. Assembly-based vulnerability of buildings and its use in performance evaluation, Earthquake Spectra 17, 291–312. Rawski, T. G., 2001. What’s happening to China’s GDP statistics?, China Economic Review 12, 347–354. Rose, A., and Liao, S.-Y., 2005. Modeling regional economic resilience to disasters: A computable general equilibrium analysis of water service disruptions, Journal of Regional Science 45, 75–112. Rose, A., Benavides, J., Chang, S., Szczesniak, P., and Lim, D., 1997. The regional economic impact of an earthquake: Direct and indirect effects of electricity lifeline disruptions, Journal of Regional Science 37, 437–458. Santos, J. R., and Haimes, Y. Y., 2004. Modeling the demand reduction input–output (I–O) inoperability due to terrorism of interconnected infrastructures, Risk Analysis 24, 1437–1451. Shaw, V. N., 1997. Urban housing reform in China, Habitat International 21, 199–212. Shirk, S. L., 1993. The Political Logic of Economic Reform in China, University of California Press, Berkeley and Oxford. Skidmore, M., and Toya, H., 2002. Do Natural Disasters Promote Long-Run Growth?, Economic Inquiry 40, 664–687. Steenge, A. E., and Bočkarjova, M., 2007. Thinking about rigidities and imbalances in postcatastrophe economies: An input–output based proposition, Economic Systems Research 19, 205–223. Strobl, E., 2012. The economic growth impact of natural disasters in developing countries: Evidence from hurricane strikes in the Central American and Caribbean regions, Journal of Development Economics 97, 130–141. Tatano, H., and Tsuchiya, S., 2008. A framework for economic loss estimation due to seismic transportation network disruption: a spatial computable general equilibrium approach, Natural Hazards 44, 253–265. Tangshan Municipal Bureau of Statistics (TBS), 1992. Fen jin de Tangshan si shi san nian 1949– 1992, China Statistics Press, Beijing (in Chinese). Tangshan Municipal Bureau of Statistics (TBS), 2008. Huihuang xin Tangshan 1978-2007, China Statistics Press, Beijing. Tangshan Municipal Bureau of Statistics (TBS), 2009. Tangshan Statistical Yearbook 2009, China Statistics Press, Beijing. Thakur, S. K., and Alvayay, J. R., 2012. Identification of regional fundamental economic structure (FES) of Chilean economy: A field of influence approach, Structural Change and Economic Dynamics 23, 92–107. Wang, Z., and Sun, D., 1996. Seismic Culture and Social Development: Revelation from New Tangshan Rise, Seismological Press, Beijing (in Chinese).

1846

WU ET AL.

Wei, S., 1994. The open door policy and China’s rapid growth: Evidence from city-level data, in Growth Theories in Light of the East Asian Experience (T. Ito and A. O. Krueger, Eds.), University of Chicago Press, Chicago. Wu, H. X., 1994. Rural to urban migration in the People’s Republic of China, The China Quarterly 139, 669–698. Wu, J., Li, N., Hallegatte, S., Shi, P., Hu, A., and Liu, X., 2012. Regional indirect economic impact evaluation of the 2008 Wenchuan earthquake, Environmental Earth Sciences 65, 161–172. Xiao, Y., 2011. Local economic impacts of natural disasters, Journal of Regional Science 51, 804–820. Yamano, N., Kajitani, Y., and Shumuta, Y., 2007. Modeling the regional economic loss of natural disasters: The search for economic hotspots, Economic Systems Research 19, 163–181. Yao, Y., 2009. The political economy of government policies toward regional inequality in China, in: Reshaping Economic Geography in East Asia (Y. Huang and A. M. Bocchi, Eds.), World Bank, Washington, D.C. Yu, S., and Su, Y., 2003. Rescue, Recovery and Reconstruction during the Aftermath of the Tangshan Earthquake, China Science and Technology Press, Beijing (in Chinese). Zou, Q., Wang, Z., Chen, F., and Wang, S., 1997. Social Recovery and Social Issues Research after the Tangshan Earthquake, Seismological Press, Beijing (in Chinese). (Received 5 September 2011; accepted 13 July 2013)