EFFECT OF SELECTED INSECTICIDE ON WHITEFLY (Bemisia tabaci) INFESTING BRINJAL PLANTS 113 A Study on Regional Specialization of China’s Agricultural Production: Recent Trends and Drivers Meilin Ma Department of Agricultural and Resource Economics, University of California, Davis, USA Sandro Steinbach Agricultural Economics Group, Swiss Federal Institute of Technology, Zurich, Switzerland Junqian Wu School of Management, Zhejiang University, Hangzhou, PR China Abstract The paper examines trends and drivers for regional specialization of China’s agricultural produc- tion for 2003-2011. We apply the Regional Specialization Index to a data set of commodities that covers 70 percent of China’s crop and livestock production. Actual prices at the provincial level are used and a seven-region framework is established that accurately reflects China’s agro- ecological characteristics. Our findings show that inequality of specialization among the Chinese regions has narrowed, with the year 2007 as a turning point. We argue that the enhanced regional specialization is due to less government intervention and more openness to international markets. Growing economic freedom, accompanied by intensified internal and external competition, has driven the regions to adjust agricultural production structures closely according to comparative advantages. Recent trends of specialization are identified and explained. Moreover, quantitative evidence to relevant agricultural policies is provided. Keywords: Regional specialization, agricultural production, China Introduction 1 Labor division and specialization are keys to economic growth (Young, 1928). Specia- lized economies employ more sophisticated technologies that result in significant output increases (Kim, 1995). In particular, region- al specialization has long been linked to re- gional development and economic growth (North, 1955; Perloff et al., 1960). With re- spect to agriculture, specialization allows a farmer to concentrate resources on specific needs (Ilbery, 1984), which increases prod- Corresponding author Name: Sandro Steinbach Email address: ssteinbach@ethz.ch uctivity (Kurosaki, 2009), rural income (Gregson, 1996; Rase and Zhang, 2009) and farmers’ welfare (Deininger and Olinto, 2001), and enhances production structure (Huffman and Evenson, 2000). On the nega- tive side, specialization could lead to soil de- terioration because the economics of scope and rotation are lost (Ilbery, 1984). Further- more, once dependent on a limited range of agricultural products, regions might become vulnerable to changes in the level of profita- bility of those products (Bowler, 1981). Yao (2004) pointed out that farmers in de- veloped countries tend to specialize and ex- plore their comparative advantages tho- roughly, which brings them more profits from trade both in domestic and global mar- Asian Journal of Agriculture and Rural Development journal homepage: http://aessweb.com/journal-detail.php?id=5005 mailto:ssteinbach@ethz.ch Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 114 kets, while developing countries are general- ly less specialized. The reasons for lower specialization in developing countries are manifold and often related to government in- terventions that cause market distortions. China’s agricultural sector offers an example where policy adjustments made by the cen- tral government have been strongly affecting the allocation of resources and influencing the production decisions of farmers through- out decades (Xu, 2001). Reviewing the his- tory of China’s agricultural development, a unique and tortuous path, from having tight central control to a much freer market, and from having agriculture sector sacrifice for industry to having industry support agricul- ture, is observed. Agricultural development in China was slow until Deng Xiaoping initiated the Chinese Economic Reform and Opening-up in 1978. Tightly restricted by the centralized planning system, regions could hardly produce ac- cording to their comparative advantages. Thus, it is widely believed that China re- mained insufficiently specialized compared with developed countries because of its dual economic structure (Liu, 2004). Agricultural development started to take off after the cen- tral government gradually loosened its con- trol over domestic production and released its constraints on the agricultural market. In 1983, the household responsibility system (HRS) was established, and long-term land- use rights were granted to farmers. By 1986, the state monopsony purchase and trade sys- tem was abolished for most agricultural commodities, except for grains and cotton, which were considered strategically impor- tant for the country. A set of further policy adjustments was initiated before China’s ac- cession to the World Trade Organization (WTO) in 2001. After the accession, impres- sive agricultural reforms were carried on. For example, in 2006, the Regulations for Farmer Professional Cooperatives (FPCs) were implemented, encouraging the estab- lishment of rural producer organizations. Additionally, agricultural taxes were re- duced since 2004 and abolished by 2007. Except for the policy adjustments mentioned above, changes on other aspects are also ob- served in China’s agricultural sector. Ac- cording to National Bureau of Statistics of China (1981-2012), the importance of agri- culture out of the entire economy has fallen from 12.6 percent in 2003 to 8.9 percent in 2011. Correspondingly, the portion of rural population fell considerably during the pe- riod. It dropped from 59.5 percent in 2003 to 48.7 percent by 2011. Noticeably, rural labor force shrank substantially in recent years, from 475.1 million in 2003 to 405.1 million in 2011. On the other hand, as shown in Fig- ure 1, the portions of agricultural production value, in the broad sense, has remained ra- ther stable. The value of crop category stays around 50 percent, while livestock accounts for 32 percent and fishery for about 10 per- cent (National Bureau of Statistics of China, 1981-2012). Being the world’s second largest economy, and the largest food producer and consumer, China’s food security, rural development, and agricultural trade issues have considera- ble global impacts. Despite the impressive achievement in the industrial, service and real estate sectors that China has achieved, its agriculture is still far from being equally successful. In light of the profound changes in agricultural markets and policies, various approaches have been chosen to figure out the reasons for the unbalanced development currently and future path for China’s agri- culture. Several scholars have tackled the questions from the perspective of agricultur- al specialization in China (Carter and Loh- mar, 2002; Tian, 2012; Xu, 2001; Yao, 2004). They agree that comparative advan- tages in the Chinese regions are not properly exploited. Carter and Lohmar (2002) were the first to provide statistical evidence for this claim. They showed that the level of specialization is far below that in developed countries. To reveal comparative advantages in the Chinese regions, the authors used the regional specialization index that was pro- posed in Krugman (1991). Our paper also takes this index as a starting point, and Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 115 moves further to make two crucial modifica- tions that improve the method to acquire a more precise picture of regional specializa- tion in China’s agricultural sector. This paper is divided into four sections. In the next section, we discuss the literature re- levant to agricultural specialization and comparative advantages of China’s regions. We continue in the third section with an overview of the methodology and the im- provements achieved through our adjust- ments. Details of our dataset are also in- cluded in this section. Our main findings are presented in the fourth section that also comprises a comprehensive discussion, which links to recent policy adjustments and market changes. The last section wraps up the paper with our core conclusions as well as remarks for future research. Figure 1: Structure of China’s agricultural production (2003-2011) Source: Data from the National Bureau of Statistics of China (1981-2011). Literature review Realizing comparative advantages allows producers in competitive markets to outper- form others and maximize profits (Carter and Li, 2002). Producers are motivated to specialize in the production of goods for which they have the lowest opportunity costs relative to competitors. Hence, region- al specialization may arise when regions ex- ploit their comparative advantages, and/or take advantage of economies of scale (Kim, 1995). Specialization is closely linked to the concept of comparative advantages that is widely used to explain production and trad- ing patterns. One famous example is the Heckscher-Ohlin model, which suggests that the comparative advantages of an entity are determined by its relative abundance of re- sources endowment. Moreover, it is impor- tant to realize that specialization has mul- tiple layers. The broadest is the country level where different countries have different ma- jor commodities in the international market. The second layer is the region level where different regions within countries specialize in the production of certain goods. Addition- ally, further specialization occurs at the farm and household level (Yao, 2004). Particularly regarding China’s agriculture, it is argued that provinces are not able to have Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 116 their comparative advantages effectively ex- ploited, leading to a relatively low degree of regional specialization (Tian, 2012). Various constraints, such as the central government’s equalization policy, stereotypes of self- sufficiency, and deficient infrastructure, have been argued to be the reasons for the insufficient specialization (Xu et al., 2008). Yet, as China goes through substantial eco- nomic changes, accompanied by a consider- able alteration in the global environment, significant adjustments in agriculture have been conducted (Carter et al., 2012). After China joined the WTO in 2001, a number of papers discussed whether the ac- cession would lead China to follow its com- parative advantages more accurately (Carter and Li, 2002). Scholars have also analyzed the endowment of input factors in China’s agriculture and their roles regarding regional productivity growth (Cho et al., 2010). These scholars applied various indices to measure the comparative advantages of agri- cultural production in different provinces (Xu, 2001; Yao, 2004; Ye, 2004). Based on the concept of regional comparative advan- tages, several papers investigated the pro- duction centralization of China’s major agri- cultural commodities (Li et al., 2012; Tian, 2012; Zheng and Cheng, 2005). None of them, however, depicted the macro picture of regional specialization sufficiently. The most thoughtful and recognized analysis of China’s regional specialization in agriculture is found in Carter and Lohmar (2002). The authors argue that the policy adjustments strongly affected regional specialization of agriculture since the 1980s. Their methodol- ogy, though, has several disadvantages that will be discussed in the following section. Methodology and data The literature provides a number of indices to measure the degree of economic speciali- zation. According to previous research, the concept of specialization takes two comple- mentary perspectives. The first perspective describes the absolute specialization. Under this approach, a subject is considered as spe- cialized if a relatively small number of in- dustries exhibit high shares of the subject’s overall employment or production value (Aiginger and Davie, 2004). Italy and its textile and food industry is a good example for such type of specialization. The perspec- tive is also called as intraregional speciali- zation, because only the internal production structure counts for the specialization degree of a region, regardless of how competitors perform. Notably, specialization measured in this way tends to show a U-shaped pattern over time (Zhang & Cheng, 2012). For this perspective, widely used indices include the Hirschman-Herfindahl Index (Herfindahl, 1950; Hirschman, 1964), the Shannon En- tropy Index (Shannon, 1948), the Diversifi- cation Index (Rodgers, 1957), and the Abso- lute Gini Index (Gini, 1921). On the other hand, specialization can be measured from a comparative perspective, called as interre- gional specialization (Fan, 2007). This ap- proach focuses on quantifying the deviation or heterogeneity of the subject’s production structure in comparison with a selected benchmark or competitors. Under this ap- proach, a revered U-shape pattern is usually found for the subject’s regions over time (Kim, 1995). Frequently used indices in- clude the Krugman Specialization Index, or the KSI (Krugman, 1991), the Index of In- equality in Productive Structure (Cuadrado- Roura et al., 1999), the Theil Index (Theil, 1967), and the Relative Gini Index (Hoover, 1936). In fact, no index could ever reflect full in- formation of the subject; but each catches a unique portion. It is crucial to choose the most appropriate index to the key question of the research. Our focus is to investigate how Chinese regions have enhanced their agricultural production structure based on comparative advantages in the course of re- cent agricultural reforms in China. Hence, it is of most importance to directly measure how dissimilarity degree has changed gradu- ally among regions after they were allowed to produce according to comparative advan- Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 117 tages in a competitive market. In other words, an index that takes the comparative perspective should be picked. Therefore, the KSI is chosen as the measure of specialization in this paper since it catches the core information to the research question. According to Palan (2010), the KSI is a superior measure for relative spe- cialization, because it is the only index that possesses the criterion of anonymity, pro- gressive transfers, merging, and bounds. In recent years, the KSI has received broad at- tention and been used in various studies of regional specialization (Kim, 1995; Krieger- Boden et al., 2008; Liang, 2004; Xian & Wen, 2006, Wolf, 2007). As mentioned, the measure has two potential references, a se- lected benchmark and all other regions. Ap- plication examples are found using both al- ternatives compared among regions (Krugman, 1991; Palan, 2010). Regarding our topic, KSI was applied to the agricultur- al sector in Kim’s paper (1995) that unfortu- nately focused on manufactory industries. In 2002, Carter and Lohmar used the KSI to measure regional specialization of China’s agriculture used it. They proposed to calcu- late the index by comparing among regions. For this paper, since we are essentially inter- ested in explicitly studying the structural distinction between one region and another in order to focus on the concept of compara- tive advantage, Carter and Lohmar’s specifi- cation is used as the final measure formula 2 . The pairwise index is defined as the sum of absolute differences over all sectors in the share of a sector in the agricultural produc- tion of a region minus the share of that sec- tor in another region’s agricultural produc- tion. The range of the index is zero to two. If the index equals zero, two regions have identical agricultural structures, while two 2 To make a sensitivity test, we also calculate the KSI using national average as the reference. The results show highly similar trend, though the ab- solute values are smaller. For details, please refer to Figure 2. regions are completely specialized if the in- dex is equal to two. We compiled a dataset that covers provincial production and price data of seventeen commodities for all Mainland China prov- inces and equivalents 3 . The dataset covers the period 2003-2011 and builds on the spe- cification presented in Carter and Lohmar (2002). We focus on the period because price data at the provincial level are not pub- lished for earlier years. Also, even national price data for earlier years are not compara- ble, because the data collection method changed in 2001. Hence, we have derived the data for 2003 to 2011 from the China Yearbook of Agricultural Price Survey (De- partment of Rural Survey, National Bureau of Statistics of China, 2003-2011), Compen- dium of Agricultural Commodity Prices and Costs (Committee of Development and Reform, 2003-2011), the Agricultural Eco- nomics Yearbooks (Ministry of Agriculture of China, 1981-2011), and China’s Statistics Yearbook (National Bureau of Statistics of China, 1981-2012). These data sources are the only trustworthy ones for regional com- parison as they are published by the central government. Indeed, other sources of data are not useful because of the inconsistent da- ta collecting methods or limited scope. Not- ably, when a province reported the produc- tion of a commodity but no price data were available for a given year, we calculate the average price for the region that the province belongs to, and use it as the province’s commodity price in that year. Less than 1 percent of the price data is calculated by this method. We included seventeen commodities from crop and livestock category. The two catego- ries account for more than 80 percent of China’s total agricultural output value which also includes fishery and forestry. To be specific, the commodities from crop catego- ry are: apples, bananas, citrus fruits, corn, 3 Municipalities (e.g. Beijing) and Autonomous Regions (e.g. Guangxi) are of the equivalent ad- ministration level as provinces in China. Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 118 cotton, flue cured tobacco, peanuts, rape- seed, rice, soybeans, sugar beets, sugarcane and wheat; and from livestock category: beef, mutton, pork and poultry. Throughout the examined period, the commodities con- tribute to over 70 percent of the total annual production value of crop and livestock cate- gory. The value of vegetables has been growing rapidly in recent years, and now occupies a considerably higher share in Chi- na’s agricultural production value. The re- gional specialization index would be even higher had we included vegetables, given the substantial production differences among provinces. Unfortunately, the complete pro- duction data at the province level are not available, and hence we have to exclude the value data of vegetables from the calcula- tion. Figure 2: Average KSI for China’s agricultural production (2003-2011) Source: Authors’ own calculation. The definition of regions is presented in Appendix. Actual prices at the province level, instead of normalized ones as earlier works did, are used. This adjustment is important because price differences are large among the com- modities (i.e. it can be as much as a 74.4- time difference) and among provinces (i.e. it can be as much as 6.1 times, and on average 2.3 times). Thus, it makes better sense to use actual prices. As shown in Figure 2, using normalized price for cross-sectional com- modities results in a considerable upward bias compared with outcomes generated from using actual price data. In other words, the bias leads to an overestimation of the degree in terms of China’s agricultural spe- cialization. Such a bias is expected, because normalized prices can enlarge the value dif- ferences of a particular commodity as the difference is based on quantity only. In fact, regions with a lower quantity tend to have a higher price, or the other around. Hence, the difference of production value is not as large as it should be if using normalized prices. Furthermore, we argue that it is more appro- priate to define agricultural regions accord- ing to natural conditions (e.g. rainfall, soil type and temperature) and historical agricul- tural production patterns. Thus, we redefine China into seven regions rather than using Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 119 the popular division of six regions, which was used by Carter and Lohmar (2002). However, there was no convincing argument for the validity of this division in the context of agriculture. For our seven-region frame- work, we rely upon the definition proposed by the Chinese Committee of National Agri- cultural Regional Planning (1981). The thir- ty-one Mainland provinces or equivalents are grouped into Northeast (NE), middle- lower division of Yellow River (YR), mid- dle-lower division of Long River (LR), South (S), Southwest (SW), Tibet-Qinghai (TQ), and Northwest (NW) Regions. An il- lustration of the six- and seven-region framework is provided in the Appendix. Notably, the seven-region framework gene- rates larger indices and clearer trends over the period studied compared with the six- region framework as shown in Figure 2 Results and discussion Regional specialization is driven by a varie- ty of factors. Theories of regional specializa- tion in agriculture classify the drivers as in- ternal and external in respect to production. Internal drivers include most fundamentally, natural endowment, technologies and labor force structure. To be specific, natural en- dowment refers to arable area, landscape, climate conditions and so on. On the other hand, external factors contain market condi- tions, infrastructure, and policies. For in- stance, market conditions refer to the open- ness of the market, the structure of the mar- ket, and the organizational structure of trad- ing entities in the market. Policies include domestic taxes, international trade regula- tions, and agricultural investments. Note that, natural endowment and labor source are rather stable over years, while technologies and policies can have considerable changes in a short period of time. Especially for Chi- na, market conditions and agricultural poli- cies have had dramatic changes during the period 2003-2011 as mentioned earlier. Thus, the following discussion is developed from an analysis of agricultural market and policies changes in recent years in order to explain the three major observations and other concerns with respect to the results. We argue that the liberalized agricultural market and supporting policies generate the major drivers, such as intensive competition and financial subsidies, for specialization enhancement in China. The major observa- tions and some concerns on the negative side are analyzed as follows. Major findings Firstly, as a whole, China’s agriculture pro- duction has become more regional specia- lized; and each region in China has become more specialized since 2003, expect for NE. Table 1 provides all average and pairwise KSI figures for the regions. The average na- tional degree of specialization, which equals the KSI divided by two, went up from 47.5 percent to 53.5 percent over nine years. We argue that the change is largely due to in- creased economic freedom both in domestic and international agricultural markets. Al- though the central government still controls the production and trade of staple foods, trade for other commodities is generally not distorted. Particularly since China joined the WTO in 2001, international agricultural markets have pushed China to move from emphasizing on self-sufficiency to balancing profit maximization. China no longer has a comparative advantage in the production of land-intensive agricultural goods (e.g. rice and wheat), while its competence in labor- intensive commodities (e.g. fruits and meats) has strengthened (Ye, 2004). Table 1: Average and pairwise KSI, using actual price data and the seven-region framework 2003 2004 2005 2006 2007 2008 2009 2010 2011 National Average 1.03 0.95 0.98 0.97 1.06 1.03 1.05 1.07 1.07 Regional Averages NE 0.95 0.87 0.84 0.86 0.87 0.89 0.87 0.89 0.91 Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 120 YR 0.96 0.88 0.93 0.92 1.01 0.97 1.00 1.03 1.02 LR 0.87 0.81 0.85 0.84 0.94 0.91 0.94 0.95 0.95 S 1.02 1.00 0.95 0.95 1.05 1.03 1.06 1.08 1.08 SW 0.86 0.81 0.82 0.81 0.88 0.86 0.88 0.89 0.91 TQ 1.50 1.34 1.43 1.35 1.51 1.44 1.46 1.49 1.48 NW 1.05 0.94 1.01 1.07 1.15 1.11 1.12 1.16 1.16 Region Pairs NE-YR 0.80 0.82 0.79 0.81 0.76 0.75 0.77 0.81 0.81 NE-LR 0.85 0.76 0.72 0.74 0.71 0.77 0.73 0.72 0.71 NE-S 0.93 0.86 0.75 0.80 0.76 0.83 0.79 0.78 0.81 NE-SW 0.79 0.65 0.60 0.62 0.58 0.62 0.57 0.59 0.67 NE-TQ 1.51 1.39 1.43 1.38 1.51 1.45 1.46 1.50 1.53 NE-NW 0.80 0.73 0.77 0.79 0.91 0.90 0.91 0.94 0.95 YR-LR 0.90 0.80 0.90 0.92 0.96 0.93 0.98 1.01 0.98 YR-S 1.08 1.08 1.03 1.08 1.17 1.11 1.19 1.23 1.18 YR-SW 0.86 0.80 0.84 0.87 0.90 0.87 0.89 0.91 0.91 YR-TQ 1.44 1.17 1.37 1.22 1.50 1.42 1.43 1.47 1.48 YR-NW 0.69 0.61 0.66 0.65 0.78 0.76 0.71 0.78 0.78 LR-S 0.37 0.41 0.38 0.43 0.48 0.46 0.49 0.52 0.54 LR-SW 0.26 0.34 0.36 0.44 0.42 0.43 0.46 0.47 0.49 LR-TQ 1.60 1.51 1.56 1.35 1.66 1.59 1.62 1.63 1.59 LR-NW 1.21 1.05 1.17 1.16 1.33 1.30 1.34 1.36 1.37 S-SW 0.53 0.57 0.51 0.56 0.54 0.54 0.57 0.60 0.60 S-TQ 1.76 1.66 1.64 1.42 1.76 1.68 1.71 1.75 1.75 S-NW 1.45 1.41 1.37 1.38 1.57 1.53 1.58 1.62 1.60 SW-TQ 1.59 1.48 1.54 1.33 1.61 1.54 1.55 1.57 1.53 SW-NW 1.10 1.00 1.06 1.07 1.24 1.18 1.20 1.22 1.25 TQ-NW 1.08 0.82 1.04 1.39 1.05 0.98 1.00 1.04 1.02 Source: Authors’ own calculation and illustration. The definition of abbreviations is given in the Appendix. Competition has motivated China to adjust its national production structure, moving from a crop-dominant to a crop-livestock- dominant agriculture with forestry and fi- shery included. For instance, before 1978, grain production occupied over 80 percent of the total production value (National Bu- reau of Statistics of China, 2005). From 2003 to 2011, the share of crops made up around 50 percent of the total value, while that of livestock commodities contributed over 30 percent. In addition to the intensi- fied competition, higher international prices as well as the growing domestic consump- tion of higher-quality and high-protein food are argued to be drivers for the structural changes (Carter et al., 2012). The high value added to livestock commodities, fruits and vegetables stimulates commercialization. Thus, an increasing number of business chains have been established to support the cycle from production to sales to services, which has resulted in more specialized agri- cultural markets. Secondly, the year 2007 is a turning point for China’s regional specialization. Since that year, the specialization degree moved up to a higher stage for most regions. Table 1 provides regional averages of specializa- tion for each region. Kernel Density Estima- tion (see Figure 3) also shows an obvious shift of specialization degree towards a higher level after 2007. The improvement did not happen overnight. Instead, several key pushing factors that were implemented in prior years gradually led to the impressive leap in 2007. These factors include the prep- aration and accession to the WTO in 2001, a jump in terms of national agricultural finan- Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 121 cial support, the reduction of agricultural taxes in 2004, and the complete nationwide abolishment of agricultural taxes in 2006. Overall, China’s switch from an agriculture- support-industry to an industry-supporting- agriculture economy greatly impacted the domestic agricultural pattern (Carter et al., 2012). The central government’s invest- ments in agriculture have grown at an annual rate of 20 to 38 percent since 2004. As a re- sult, China spent 9.6 percent of its entire budget on agriculture in 2011 (Ministry of Agriculture of China, 1981-2011). This cor- responds with a nearly six-fold increase in agricultural investments in the investigated period. The financial support includes subsi- dies on agricultural production, the supply of high-quality seeds and fertilizers, the con- struction of infrastructure, and loans pro- vided by rural credit systems. Consequently, farming entities have less financial burden and enjoy improved producing conditions. Figure 3: Kernel density estimation for regional average KSI Source: Authors’ own calculation. The Kernel Density Estimation (KDE) is a non-parametric way to estimate the probability density function of a random variable. KDE is a fundamental data-smoothing problem where inferences about the population are made, based on a finite data sample. Thirdly, we notice that the most similar re- gion pairs distinguish themselves most rapid- ly, leading to substantial increases in their own overall specialization degree (see Table 1). Actually, the coefficient of variance for regional average KSI also drops from 0.213 in 2003 to 0.189 in 2011. 4 Since the coeffi- cient of variance measures the deviation of 4 The coefficient of variance is a normalized measure of dispersion of a probability distri- bution or frequency distribution. the region’s specialization level from the na- tional average as the standard deviation di- vided by the mean, the decrease shows that the gap of specialization levels among re- gions has narrowed. In other words, speciali- zation has increased in every region, and the inequality of regional specialization degrees has decreased. To be specific, the least distinguished region pair is LR and SW in 2003, whose index went up from 0.26 to 0.49 after nine years. An ana- Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 122 logous trend is observed for the LR and S pair. Looking into details, it is found that in 2003, the production patterns of staple food and meat in LR and SW were highly similar. Staple food accounted for 36 percent and meat 47 percent of LR’s total production val- ue, while the shares were 37 percent and 47 percent for SW, respectively. In the investi- gated period, LR gradually specialized in the production of staple food, which is consi- dered as more land-intensive. The region’s comparative advantages include better irriga- tion systems and flatlands, which are more suitable for rice production. Moreover, the coastal provinces in LR have prosperous economies, leading to a booming food de- mand. Additionally, the region has the high- est machinery level that allows for lower pro- ductions costs, and potentials for economies of scale. On the other hand, SW shifted from being staple-food-dominant to focusing on meat production, which is generally more la- bor-intensive. The comparative advantages of SW lie in their lower labor and feeding costs. Therefore, in 2011, the production of staple food accounted for 44 percent of total pro- duction value in LR and only 29 percent in SW, while the production of meat contributed 40 percent in LR but 53 percent in SW. A similar pattern is observed for tobacco pro- duction in LR and SW. In fact crop and livestock category in the two regions have traditionally had similar patterns due to their somewhat similar natural condi- tions, while noticeable differences are found only in forestry and fishery. To be specific, with rich water resources, LR has a much stronger fishery than SW, while the moun- tainous landscape of SW contributes to its strength in forestry. Knowing that the distinc- tion of production structures cannot be brought only by natural condition changes given the short time span, we argue that the increasing interregional and international competition is driving the regions to focus on their comparative advantages. Thus, the re- gions ended up differentiating their produc- tion structures. Our argument is supported by China’s inter- national trade data both at the national and regional levels. China’s agricultural trade growth was substantial in the investigated pe- riod. Imports grew 397 percent, reaching al- most US$ 109 billion in 2011, while exports grew 179.3 percent, equal to US$ 58.1 billion (United Nations, 2013). The export growth is driven by horticultural (i.e. vegetables and fruits) trade, which has grown by 316.7 per- cent, with its share in total agricultural trade increasing from 9.8 percent in 2003 to 14.8 percent in 2011. Export growth rates regard- ing other commodities are substantially smaller. The fact indicates a shift towards ex- port trade in high-value products. On the im- port side, a similar trend is observed, but rates are significantly larger. These changes reflect the comparative advantages in China’s re- gions have regarding international markets on average. Out of all the seven regions, LR, S, and SW record the fastest trade growth rates, increasing their aggregated agricultural trade by roughly 300 percent in the period 2003- 2010 (Ministry of Agriculture of China, 1981-2011). For example, total exports and imports for S reached a historic height of US$ 27.1 billion in 2010. The fact indicates that the three regions are influenced by the inten- sified international competition the most, which gives the producers strong incentives to focus on their comparative advantages. Meanwhile, it is important to notice that al- though international trade in China is grow- ing, domestic trade still dominates. In 2010, only 1.8 percent of China’s total agricultural production value was from international trade (Ministry of Agriculture of China, 1981- 2011). This implies that specialization is still mainly pushed forward by interregional com- petition. We find that the number of FPCs and the total assets of township agribusi- nesses have increased impressively recently. In 2006, the Regulations for Farmer Coopera- tives were implemented. Thus, legal assur- ance was provided that motivated the devel- opment of FPCs, contributing to an increased ability of small-scale farmers to benefit from economies of scale (Chen, 2009). The num- Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 123 ber of FPCs in China reached 521,700 by 2011 (Xu, 2012). Finally, LR, S and SW are pioneering the ex- pansion of FPCs and township agribusiness. For example, by 2010, the total assets owned by township agribusinesses in China already doubled in comparison with the record of 2003, and that of LR reached a record of US$ 17.2 billion (National Bureau of Statistics of China, 1981-2012). Clearly, even though households still mainly conduct farming, the boom of FPCs and agribusinesses indicate that the value-chain integration is streng- thened, both horizontally and vertically. Larger groups of producers have more bar- gaining power and are more capable of ac- quiring technology, market access and econ- omies of scale. Having more economic free- dom and closer market relations causes orga- nized producers to compete more intensively, which forces producers to focus on compara- tive advantages, resulting in regional specia- lization (Yang, 2011). Impediments and opportunities for future regional specialization In spite of the observed improvements, Chi- na’s regional specialization lags behind that of developed countries, implying huge agro- economic potentials from further specializa- tion. Taking the USA as an example, the na- tional specialization index was 62.5 percent in 1987 (Kim, 1995). This is substantially higher than that of China in 2011 (53.5 per- cent). There are a number of key factors that negatively affect agricultural specialization in China. Firstly, the central government still controls most of the staple food production and trade, and sets the food self-sufficiency require- ments. In terms of global markets, state- owned companies dominate trading relations. In addition, it is a major obstacle for further specialization that agribusinesses are not granted the full authority to participate in in- ternational markets. Basically, they are merged as part of state-owned companies through contracts, and the companies will then trade for them. For instance, the most powerful Chinese food company is Cereal, Oil and Foodstuffs Importing and Exporting Corporation (COFCO). On the positive side, it is safer for small-scale producers to special- ize when revenues are guaranteed by contract farming. On the negative side, however, far- mers may not be able to produce according to their comparative advantages. Furthermore, though Chinese farmers are now free to lease and transfer their land, selling land is not yet legal according to China’s Land Law. Thus, overall, land fragmentation is still serious (Xu et al., 2008). The average arable land per ca- pita remained at only 0.153 hectare by 2011 (National Bureau of Statistics of China, 2011). The uncertainty of long-term land use rights leads to low land integration, deters ru- ral households from investing in land and limits production specializing. It is noticeable that the Chinese government has released strong signals of modernizing agriculture and enhancing agricultural specia- lization. For example, as described in the Na- tional Plan for Modern Agricultural Devel- opment 2011-2015, Beijing announces the es- tablishment of national production centers and production-belts (The Central Committee of the Communist Party of China and State Council, 2012). The plan includes bring in more investments to agricultural research and development, the improvement of agricultural infrastructure, increasing rural income, and enhancing industrialization of agricultural production. Such policies are expected to mo- tivate specialization of agricultural produc- tion in the coming years. Conclusion and further thoughts The scope of this paper is to explore the trends and drivers of regional specialization in China’s agriculture. Krugman’s regional specialization index is employed and mod- ified. First, instead of normalizing prices as did in earlier studies, we use actual price data on provincial level. In addition, we argue that it is more appropriate to define agricultural regions in China based on their natural condi- tions and agricultural production patterns. Thus, our new seven-region framework draws Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 124 a more precise picture of regional specializa- tion. We apply the augmented index to a da- tabase, which covers more than 70 percent of China’s crop and livestock production value for the period 2003-2011. Our findings show that Chinese agriculture has moved to a high- er level of regional specialization, especially after 2007. We have identified several drivers of these changes. Firstly, the considerable in- crease in agricultural specialization has re- sulted from less government intervention, which has led to more economic freedom, prompting regions to adjust their production patterns. Secondly, the liberalized domestic market and increased exposure to internation- al markets have resulted in intensive internal and external competition, forcing regions to produce based on comparative advantages. Thirdly, the commercialization of agriculture, accompanied by the horizontal and vertical integration of the value-chain, motivated a structural change in China’s agriculture. The change is also reflected by the prosperity of FPCs and agribusinesses during the period It is worth mentioning that the paper has several limitations that hold good potentials for fu- ture research. Firstly, the time span of the study is relative short. A longer period is de- sired to figure out a general developing pat- tern of regional specialization in China. Se- condly, value data of vegetables and some fruits are lacked with current data source. Adding in those data can be particularly help- ful, given that vegetables and fruits are in- creasingly more important of China’s agricul- ture. In addition, since the paper is mostly de- scriptive, it requires more rigorous proof to formulize the structural changes or thresholds of regional specialization over time. Lastly, as we discussed earlier, there are other indic- es that can be used to study specialization from different perspectives. Especially given the multi-dimensions of specialization, it is promising to use other indices and look for more insight of regional specialization with alternative approaches. Regional specializa- tion goes hand in hand with higher agricul- tural production values. Hence, it is crucial for China to increase the attractiveness in farming to ensure a sufficient level of farm labor in rural areas, especially as rural labor flooding to cities. Besides, since the 1980s, China’s industrial, service and real estate sec- tors have grown rapidly, while agriculture has been left behind, resulting in a great urban- rural development gap. Closing up the ever- widening gap is a historical challenge for China. To tackle the task, Beijing has already announced a series of policies that enhance agricultural infrastructure, improve agricul- tural research and development, and encour- age deeper regional specialization in agricul- tural production. These policies promise a higher level of regional specialization in agri- culture, with their implications far-reaching. All in all, China has substantial growth poten- tials in agriculture that promise a more ba- lanced and sustainable development. Acknowledgement The authors sincerely thank Lina Yamashita, Li Gucheng, Alex F. McCalla, and Stephen Boucher for their comments on the early ver- sions of this paper. The research was sup- ported by the National Natural Science Foun- dation of China (No. 71273103) and the Pro- gram for New Century Excellent Talents in University of China (No. NCET-11-0647). References Aiginger, K., & St. Davies (2004). Industrial specialization and geographic concen- tration: two sides of the same coin? Not for the European Union. Journal of Applied Economics, 12, 231-248. Bowler, I. R. (1981). Regional specialization inthe agricultural industry. Journal of Agricultural Economics, 32, 43–54. Carter, C. A., & Li, X. (2002). Implications of world trade organization accession for China's agricultural trade patterns. The Australian Journal of Agricultural and Resource Economics, 46(2), 193- 207. Carter, C. A., & Lohmar, B. (2002). Regional specialization of China's agricultural production. American Journal of Agri- cultural Economics, 84(3), 749-753. Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 125 Carter, C. A., Zhong, F., & Zhu, J. (2012). Advances in Chinese agriculture and its global implications. Applied Eco- nomic Perspectives and Policy, 34(1), 1-36. Chen, X. (2009). Review of China's agricul- tural and rural development: policy changes and current issues. China Agricultural Economic Review, 1(2), 121-35. Cho, S. H., Chen, Z., & Poudyal, N. C. (2010). Spatial structure of agricultural production in China. Applied Econom- ics, 96(1), 2031-2040. Committee of Development and Reform (2003-2011). Compendium of Agricul- tural Commodity Prices and Costs. China Statistics Press, Beijing. Committee of National Agricultural Regional Planning (1981). Comprehensive Re- gional Planning of China's Agriculture. Agriculture Press, Beijing. Cuadrado-Roura, J. R., Garcia-Greciano, B., & Raymond, J. L. (1999). Regional convergence in productivity and pro- ductive structure: the Spanish case. In- ternational Regional Science Review, 22, 35-53. Deininger, K., & Olinto, P. (2001). Rural nonfarm employment and income Di- versification in Colombia. World De- velopment, 29(3), 455–465. Department of Rural Survey, National Bureau of Statistics of China (2003-2011). China Yearbook of Agricultural Price Survey. China Statistics Press, Beijing. Fan, F. (2007). Measuring regional speciali- zation. Journal of Economic Research, 9, 71-83. Gini, C. (1921). Measurement of inequality of incomes. The Economic Journal, 31, 124 -126. Gregson, M. E. (1996). Long-term trends in agricultural specialization in the Unit- ed States: some preliminary results. Agricultural History, 70(1), 90-101. Herfindahl, O. C. (1950). Concentration in the steel industry, PhD-Dissertation, Columbia University, New York. Hirschman, A. O. (1964). The paternity of an index. The American Economic Re- view, 54, 761-762. Hoover, E. M. (1936). The Measurement of industrial localization. The Review of Economics and Statistics, 18, 162-171. Huffman, W. E., & Evenson, R. E. (2000). Structural and productivity change in US agriculture. Agricultural Econom- ics, 24(2), 127-147. Ilbery, B. W. (1984). Agricultural specializa- tion and farmer decision behaviour: a case study of hop farming in the west midlands. Tijdschrift Voor Econo- mische En Sociale Geografie, 75, 329– 334. Kim, S. (1995). Expansion of markets and the geographic distribution of economic activities: the trends in U. S. regional manufacturing structure, 1860-1987. The Quarterly Journal of Economics, 110(4), 881-908. Krieger-Boden, C., Morgenroth, E. L. and Pe- trakos, G. (2008). The Impact of Euro- pean Integration on Regional Struc- tural Change and Cohesion (1st ed.) Routledge, London. Krugman, P. (1991). Geography and trade, MIT Press, London. Kurosaki, T. (2009). Specialization and di- versification in agricultural transfor- mation: the case of West Punjab, 1903- 92. American Journal of Agricultural Economics, 85(2), 372-386. Li, E., Pang, A., & Zhu, J. (2012). Analysis of the evolution path and mechanism of China's agricultural agglomeration and geographic pattern. Geographical Research, 31(5), 885-898. Liang, Q. (2004). The division of labor in manufacture, regional specialization in China with an international compari- son. Global Economics, 12, 32-40. Liu, M. (2004). Division impeded and institu- tionalized poverty trap. Issues in Agri- cultural Economy, 2, 53-57. Ministry of Agriculture of China (1981-2011) China Agricultural Yearbook, China Agriculture Press, Beijing. Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 126 National Bureau of Statistics of China (2011). China household survey yearbook, china statistics press, Beijing. National Bureau of Statistics of China (1981- 2012). China statistics yearbook, Chi- na Agriculture Press, Beijing. National Bureau of Statistics of China (2005). China Compendium of Statistics 1949- 2004, China Statistics Press, Beijing. National Bureau of Statistics of China (2004- 2012). China rural statistics year book, China Statistics Press, Beijing. National Bureau of Statistics of China (2004- 2012). China yearbook of agricultural price survey, China statistics press, Beijing. North, D. C. (1955). Location theory and re- gional economic growth. Journal of Political Economy, 63(3), 243-258. Palan, N. (2010). Measurement of specializa- tion – the choice of indices. FIW Working Paper, No. 62. Perloff, H., Dunn, E., Lampard, E., & Muth, R. (1960). Regions, resources, and Economic Growth, John Hopkins Press, Baltimore. Rase, A. N., & Zhang, X. (2009). China’s booming livestock industry: household income, specialization, and exit. Agri- cultural Economics, 40(6), 603-616. Rodgers, A. (1957). Some aspects of indus- trialh diversification in the United States. Economic Geography, 33: 16- 30. Shannon, C. E. (1948). A mathematical theory of communication. The Bell System Technical Journal, 27(379- 423), 623-656. The Central Committee of the Communist Party of China and State Council (2012). National Plan for Modern Agricultural Development 2011-2015, The Central Committee of the Com- munist Party of China and State Coun- cil, Beijing. Theil, H. (1967). Economics and Information Theory, North Holland, Amsterdam. Tian, W. (2012). Study on the Rules of the Spatial Structural Evolution of Agri- cultural Development in China, PhD- Dissertation, Southwest University, Chongqing. United Nations (2013). International Mer- chandise Trade Statistics, United Na- tions Statistical Division, New York, available at http://comtrade.un.org/db/. Wolf, N. (2007). Endowments vs. market po- tential: what explains the relocation of industry after the Polish reunification in 1918? Explorations in Economic History, 44, 22-42. Xian, G., & Wen, D. (2006). FDI, regional specialization and industrial clusters. World of Management, 12, 18-31. Xu, Q., Tian, S., Xu, Z., & Shao, T. (2008). Rural land system, land fragmentation and farmer's income inequality. Eco- nomics Studies, 2, 83-105. Xu, X. (2012). The situation of farmer coop- eratives of China. China Industry and Commerce News, 12/17/2012, http://www.cicn.com.cn/content/2012- 12/17/content_121153.htm. Xu, Z. (2001). Comparative Advantage and Production Structure Adjustment in Chinese Agriculture, PhD- Dissertation, Nanjing Agricultural University, Nanjing. Yang, D. (2011). Research on Farmer Coop- erative Economic Organization Pro- moting Agriculture Labor Division and Specialization Development – Based on Chinese Family Production Mode, PhD Dissertation, Southwest Universi- ty, Chongqing. Yao, S. (2004). Specialization and Agricul- tural Development: Theoretic Re- search and a Case Study of China, PhD-Dissertation, Southwest Univer- sity of Finance and Economics, Chengdu. Ye, C. (2004). Comparative Advantage and Structure Adjustment in China's Crop Production. PhD-Dissertation, Nanjing Agricultural University, Nanjing. Young, A. A. (1928). Increasing returns and economic progress. The Economic Journal, 38(152), 527-542. Zhang, J., & Cheng, W. (2012). The U-shape pattern in regional industrial speciali- http://comtrade.un.org/db/ http://www.cicn.com.cn/content/2012-12/17/content_121153.htm http://www.cicn.com.cn/content/2012-12/17/content_121153.htm Asian Journal of Agriculture and Rural Development, 4(2)2014: 113-127 127 zation in China. Chinese Social Sciences, 1: 76-97. Zheng, F., & Cheng, Y. (2005). From agricul- tural industrialization to agricultural industrial district: study on the feasibil- ity of industrialization of competitive agriculture. World of Management, 7, 64-73. Appendix Definition of China’s regions (six- and seven-region framework). Northeast (NE): Heilong- jiang, Jilin, and Liaoning; North (N): Beijing, Hebei, Henan, Shandong, Shanxi, and Tianjin; Northwest (NW): Gansu, Inner Mongo- lia, Ningxia, Qinghai, Shaanxi, Tibet, and Xinjiang; East (E): Anhui, Jiangsu, Shanghai, and Zhejiang; Central (C): Chongqing, Hu- bei, Hunan, Jiangxi, and Si- chuan; and South (S): Fujian, Hainan, Guangdong, Gua- ngxi, Guizhou, and Yunnan. Northeast (NE): Heilong- jiang, Jilin, and Liaoning; Northwest (NW): Gansu, In- ner Mongolia, Ningxia, and Xinjiang; middle-lower divi- sion of Yellow River (YR): Beijing, Hebei, Henan, Shaanxi, Shandong, Shanxi, and Tianjin; middle-lower division of Long River (LR): Anhui, Hubei, Hunan, Jiang- xi, Jiangsu, Shanghai, and Zhejiang; South (S): Fujian, Guangdong, Guangxi, and Hainan; Southwest (SW): Chongqing, Guizhou, Si- chuan, and Yunnan; and Ti- bet-Qinghai (TQ): Tibet and Qinghai. Source: Authors’ own illustration