Systematic Optimization of China’s Manufacturing Industrial Structure
ShiDan(史丹)andZhangCheng(张成)
Abstract:
Using China’s two-digit manufacturing sectors as samples, this paper first analyzes China’s output structure optimization objectives and energy conservation and emissions abatement potentials in 2015, then examines various factor inputs’ matching, and estimates their capacity utilization status, focusing on capital stock factor. Results of our study suggest that: (1) China’s manufacturing output structure has great potentials of optimization to reduce energy intensity and carbon intensity by 18.08% and 17.42% respectively over the original values; (2) to reduce factor mismatch, various supporting input factors need to be introduced after manufacturing output structure optimization. The level of capital stock, in particular, requires a substantial change; (3) China’s manufacturing capacity utilization (56.14%) in 2015 was far below its average level (73.27%) in the mid and late stage of the 11th Five-Year Plan period (2008-2010). The low capacity utilization was attributable to economic slowdown and investment inertia. After input factor matching, capacity utilization may rise to the latter level.
Keywords:
output structure, factor structure, overcapacity, energy conservation and emissions abatement JEL Classification Codes: O21; Q01; Q56 DOI: 1 0.19602/j .chinaeconomist.2018.11.01
1. Introduction
Over the past four decades, China’s manufacturing industry has contributed a significant share to its rapid economic growth, job creation and the “China miracle” through rapid industrial structure evolution. Today, China boasts the largest manufacturing industry in the world, with significant improvements in its national power and international competitiveness. In its current stage, China has to simultaneously deal with the slowdown in economic growth, make difficult structural adjustments, and absorb the effects of previous stimulus policies. Its manufacturing industry is faced with the dilemma between growth stability and structural adjustment, as well as competitive pressures from both developed countries and emerging economies. While low-cost advantages diminished, new competitive edges are yet to develop.
These challenges cast shadow on the future outlook of China’s manufacturing industry. There has been a great deal of interest among researchers regarding how to make China’s manufacturing industrial structure more advanced and reasonable in order to promote the quality and efficiency of economic development.
The question to be discussed in this paper is how to adjust China’s manufacturing industrial structure1. Existing studies attempt to answer relevant questions in this field. Based on a scientific evaluation of the historical role of China’s output structure evolution (Liu and Zhang, 2008; Zhang, 2010), studies simulate the directions of industrial structure optimization and its counterfactual effects (Wang and Xiang, 2014; Zhu et al., 2014; Zhang and Zhao, 2015). In optimizing output structure, existing studies introduce such factors as energy conservation and emissions abatement, employment security and industrial coordination. However, these studies are confined to an isolated economy’s perspective without using relevant information of an open economy. Although existing studies discuss output structure optimization and production factor allocation (Ngai and Pissarides., 2007; Yuan and Jie, 2011; Benhima, 2013; Dong, 2015), the two issues are not properly integrated. In optimizing output structure, almost all studies only provide the desirable output levels of various sectors without revealing the extent to which capital, labor and other inputs should be adjusted accordingly. Factor structure matching analysis, which is absent in these studies, can be introduced in the industrial structure optimization.
To facilitate theoretical research and provide policy recommendations, this paper conducts a systematic industrial structure optimization using China’s two-digit manufacturing sectors as samples. In addition to analyzing output structure optimization objectives and energy conservation and emissions abatement potentials, this paper also investigates questions of input factor matching and capital stock capacity utilization. This paper has the following contributions: (1) In optimizing manufacturing output structure, this paper takes into account other factors in a more comprehensive and scientific manner, including demand and supply information. In particular, this paper considers demand-side import/export potentials and supply-side technology contribution, which are seldom mentioned in existing studies; (2) unlike existing studies which separately examine the structural optimizations of output and factor, this paper integrates the analysis of structural optimization with factor input matching; (3) in investigating factor structure matching, this paper follows an approach of succession and criticality. In matching factor structure by extracting historical information, we offer a deeper analysis of capital factor allocation to address the potential problem of capital factor overcapacity.
2. Model and Research Methodology
With respect to research methodology, this paper adopts the following steps: Step 1: Non-linear Programming is employed to optimize China’s manufacturing output structure of 2015 from an energy conservation and emissions abatement perspective, taking into account factors such as employment security, industrial equilibrium, import/export potentials and technology contribution. Step 2: After obtaining a non-linear relationship between factor input and economic output, trans-log production function model is employed to match an appropriate factor pattern for optimized output structure. Step 3: Data Envelopement Method (DEA) is employed to estimate capacity utilizations before and after optimization, focusing on capital stock factor.
2.1 Creation of Non-linear Programming Model
Based on above theoretical discussions, this paper assumes that total energy consumption and CO2 emissions cannot exceed ceilings under the conditions of employment security, input-output equilibrium, final domestic consumption potentials, import/export potentials, as well as technology contribution. In order to minimize overall national resource and environmental intensity (weighted energy and carbon intensities), the following Non- linear Programming is specified to seek manufacturing industrial structure optimization2: i(j), t and b in equations (1) through (11) respectively denote sector3 ( i= 1,2… m；j= 1,2… m+n), year and energy type ( b= 1,2… k), and t0 denotes a year before year t. * denotes the result after optimization. TP is resource and environmental intensity. EP, CP and LP are energy intensity, carbon intensity and labor intensity respectively. and are the weight ratio coefficients of EP and CP respectively. Y, E, C, L, XF, IM, EX and RT are output, energy, CO2, labor, other consumption4, import value, export value and technology contribution. and are changes in import value and export value. is direct consumption coefficient, and is change in direct consumption coefficient. is change in other consumption. is change in national total labor.
Equation (1) is objective function, i.e. seeking the minimization of overall national resource and environmental intensity. Equations (2) through (11) are constraints. Among them, equations (2) through (4) create relationships between output and energy consumption, and CO2 emissions and labor quantities of various sectors. Equation (5) restrains sector outputs from an inter-sector equilibrium and import/export
perspective5. Equation (6) ensures from a technology contribution perspective that after optimization, the total contribution of various sectors’ technology levels to output is at least no less than that of their original level before optimization. Equations (7) through (9) provide constraints on total energy consumption, CO2 emissions and employment security. Equations (10) and (11) respectively provide the methods for calculating national energy intensity and carbon intensity.
2.2 Creation of Trans-Log Production Function Model
Under relevant assumptions and constraints, this paper is able to obtain the output size of China’s two- digit manufacturing sectors after optimization. But a new question is how various sectors should make use of input factors to efficiently provide desirable output and reduce factor mismatch. For this purpose, historical data can be employed to estimate the non-linear relationship between input factors and output, and calculate a reasonable factor allocation pattern according to the needs of output.
In estimating the non- linear relationship between factor input and output, this paper employs stochastic frontier analysis (SFA) method since this method is able to not only decompose technical efficiency values from productivity but control for the disturbance arising from stochastic error term, so as to more accurately depict the relationship of substitution or supplement between factor inputs, as well as the non-linear relationship between factor inputs and outputs. Based on Battese and Coelli’s (1995) SFA model and referencing existing literature, this paper adopts a function form including capital ( K), labor ( L), intermediate product input ( M) and technology level ( T). In order to examine factor input’s marginal output and elasticity in more detail, this paper specifies production function in the trans-log form, whose specific form is as follows:
Where, β is parameter to be estimated; U is output inefficiency, which conforms to iid and denotes output loss caused by differences in the internal management levels of decision-making units. V is stochastic deviation term, which satisfies iid , and denotes luck’s stochastic impact on output.
Once the size of desirable output of each manufacturing sector is obtained, the labor quantity and intermediate input quantity which sectors need to absorb can be estimated. Then, equation (12) can be used to estimate the appropriate size of capital stock.
2.3 Application of Data Envelopement Method
This paper employs the input-oriented non- discretionary variable model ( NDSC) created by Cooper et al. ( 2004) with constant return to scale. This model is able to extract information of discretionary variable (capital stock) and non-discretionary variables (labor and intermediate input), and focus on analyzing input efficiency of discretionary variable under the condition of specifying the non-discretionary variables as constants. Due to limit of length, this model will not be described in detail.
After calculating capital redundancy ( ), capital utilization ( ) can be obtained using the following equation (13):
3. Model Creation and Data Explanation
This paper employs 2003-2015 panel data of manufacturing two-digit sectors of 30 provincial-level regions (excluding Tibet, Hong Kong, Macao and Taiwan), and data is arranged and calculated according to province-specific statistical yearbooks, DRCnet database and the China Statistical Application Support System. In order to exclude the impact of price factor, all price-related data in this paper is adjusted to the 2000 price level according to relevant price index or growth index. Given the differences in the 2002 and 2011 editions of national economic sector classification, this paper conducts necessary data splits and merges to form 29 manufacturing sectors6.
In empirical analysis, the following variables are created: (1) output ( Y): Actual aggregate industrial value after adjusting for the ex-factory prices of industrial goods; (2) labor input ( L): Denoted by yearend total employment; (3) capital input ( K): Calculated using perpetual inventory method with equation
. In calculation, the method provided by Dong et al. (2015) is employed. Where, It is new investment value, denoted by the difference between the original prices of fixed assets of two adjacent years; Pt is the price index of investment goods, denoted by fixed asset investment price index; depreciation rate is measured by the mean value7 of estimated depreciation rates of various years; base-period capital stock is expressed by the difference between original price of fixed assets and cumulative depreciation in 2000. (4) Intermediate product input ( M): After subtracting industrial valueadded and payable tax increase from gross industrial output value, the result is divided by raw material purchase price index. Industrial value-added data of 2001-2007 for two-digit industrial sectors is from China Industrial Statistical Yearbook, and the data of 2008-2015 is expressed by the product between current-year gross industrial output value and average industrial value-added ratios of 2003-2007. (5) Technology progress ( T): If technology progress needs to be introduced into trans-log production function model, it should be depicted using time spans 1-13. (6) Energy consumption ( E): End-user energy consumption adopted in our calculation includes raw coal, cleaned coal, other washed coal, briquette, coke, coke oven gas, other coal gas, crude oil, petroleum, coal oil, diesel, fuel oil, liquefied petroleum gas, refinery dry gas, natural gas, other petroleum products, other coke products, heat power and electric power, and is converted into standard coal equivalent using standard coal conversion coefficient provided by the National Bureau of Statistics (NBS). (7) CO2 emissions ( C): CO2 emission factor of conventional fossil energy is subject to data provided by IPCC (2006). The CO2 emission factor of electric power as secondary energy is from the national benchmark data provided by the National Center for Climate Change Strategy and International Cooperation (NCSC). It is assumed that all heat power is generated from raw coal combustion and converted according to raw coal’s emission factor. (8) Energy intensity ( EP), carbon intensity ( CP) and labor intensity ( LP): Denoted by the ratio of energy consumption, CO2 emissions and labor quantity to industrial gross output. (9) Import ( IM), export ( EX), direct consumption coefficient ( ) and other consumption ( XF): Based on China’s input-output tables of 2012, data of relevant sectors is calculated in combination. Where, other consumption is measured
by the sum between the total indirect consumption of manufacturing products by all sectors other than a country’s manufacturing sectors and the final consumption by households, government and capital. (10) Energy intensity’s weighting coefficient ( ): The weighting coefficients of energy intensity and carbon intensity are both 0.5. (11) Contribution of technology level ( TP): Ratio between output growth induced by technology level and total output7. (12) Change in national total workforce ( ): The mean value of absolute values of change in national total workforce in recent three years is the upper and lower restricted line. (13) Change in import value ( ), change in export value ( ), change in direct consumption coefficient ( ) and change in other consumption ( ). According to existing data, difference method is employed to estimate the change in 2015 relative to 2012.
4. Empirical Results and Analysis
4.1 Manufacturing Industrial Structure Optimization
Based on the Non-linear Programming provided in the above section, this paper estimates the size of desirable output for various manufacturing sectors in 2015 and their energy consumptions and CO2 emissions from an energy conservation and emissions abatement perspective, i.e. minimization of resource and environmental intensity.
(1) Potential effect after manufacturing output structure optimization
In 2015, China’s manufacturing gross output value was 89,856.41 billion yuan, and optimized gross output value may increase to 93,883.30 billion yuan, up 4.48%. Meanwhile, the energy conservation and emissions abatement effects are favorable: Total energy consumption may reduce from 2,110.17 million tce to 1,806.19 million tce (down 14.41%), and total CO2 emissions may reduce from 6,096.11 million tons to 5,260.01 million tons (down 13.72%). In this manner, resource and environmental intensity reduced from 4,566 tons/100 million yuan to 3,763 tons/100 million yuan (down 17.59%). Specifically, energy intensity reduces from 2,348 tce/100 million yuan to 1,924 tce/100 million yuan (down 18.08%), and carbon intensity reduces from 6,784 tons/100 million yuan to 5,603 tons/100 million yuan (down 17.42%).
(2) Manufacturing sector output structure optimization and analysis
This paper drafts the following Figure 2 to more clearly reveal the direction and degree of output size adjustment after a comparison between optimized values for manufacturing sectors in 2015 and their original values in 2015 and 2010. Meanwhile, this paper introduces the following six categories to give a clear picture of sectors’ adjustments and patterns: (1) strong absolute production increase, (2) weak absolute production increase, (3) relative production increase, (4) absolute production reduction, (5) strong relative production reduction and (6) weak relative production reduction. Refer to Table 1 for the criteria of classification for each category. Strong absolute production increase means that the sector’s output size is not only greater than the original value of 2015 but also greater than the average growth rate of the optimized value of 2015 relative to 2010 (66.57%, referred to as “benchmark growth rate”). One may only need to observe the “●” and “▲” labels of various sectors in Figure 2. If they are all above their critical lines of 0% and 66.57%, the sector is of absolute production increase. After an observation, we know that the following nine sectors meet this criterion, including manufacture of medicines, manufacture of special purpose machinery, manufacture of electrical machinery and equipment, manufacture of communication equipment, computers and other electronic equipment and recycling and disposal of waste.
Strong relative production reduction means that despite the increase of a sector’s output size over 2010, it is smaller than benchmark growth rate and the original value of 2015, i.e. “●” and “▲” symbols should be smaller than 0% and in the range of [0%,66.57%]. Obviously, 17 sectors are of this category, including processing of food from agricultural products, manufacture of foods, manufacture of paper and paper products, manufacture of rubber and ferrous metal smelting and pressing.
Weak absolute production increase means that a sector’s output size is greater than the original value of 2015, and has some growth compared with 2010, but is smaller than benchmark growth rate, i.e. the sector’s “●” and “▲” symbols should be higher than 0% and in the range of [0%,66.57%]. Sectors of this category include transport equipment manufacturing.
Weak relative production reduction means that a sector’s output size is smaller than original value of 2015, but is greater than benchmark growth rate. If a sector’s “●” symbol is smaller than 0% critical line, but “▲” symbol is higher than 66.57% critical line in Figure 2, this sector is of weak relative production reduction. Only manufacture of beverage and manufacture of cultural, educational, fine arts, sports and entertainment goods are of this category. This indicates that even compared with the original values of 2015, the output size of these sectors should be appropriately reduced, but compared with the manufacturing industry’s overall benchmark growth rate, they are still higher than average level. No sector can be classified into the other categories.
Obviously, the nine manufacturing sectors of strong absolute production increase include not only high-tech advanced manufacturing and high-end equipment manufacturing, but the promising internet industry, as well as “venous industry” which tends to be overlooked. Without doubt, advanced manufacturing, high-end equipment manufacturing and internet sectors, including “internet+” sectors, are key to the success of China’s “Industry 4.0” and “Made in China 2025” roadmap. Despite its limited share of output, recycling and disposal of waste as a “venous industry” enjoys superior growth
momentum after optimization among sectors of strong absolute production increase: Its original output value needs to be increased from 327.99 billion yuan to 591.76 billion yuan, up 80.42%.
Sectors of strong relative production reduction should be properly understood. Gross output reductions of these sectors with high resource and environmental intensities represent an overall optimization based on an input- output framework under the condition of satisfying consumption, investment and all sectors’ demand for intermediate inputs, import/export restrictions and technology contributions. In order to minimize overall resource and environmental intensity and avoid overcapacity, these sectors should give way to sectors of strong absolute production increase to some extent. Of course, this does not mean that each firm should simply cut production. Rather, more resource-consuming and polluting firms in various sectors should be closed or change production through a survival-of-thefittest process to meet the gross output size targets of various sectors. Firms with relative comparative advantages should expand to achieve economies of scale and economies of scope. As for sectors of weak relative production reduction, their basic conditions are similar to those of sectors of strong relative production decrease. The only difference is that while their output value needs to reduce to some extent, their optimized growth rates are still higher than benchmark growth rate.
4.2 Factor Structure Matching for Optimal Manufacturing Output Structure
(1) Estimation result and analysis of trans-log production function model
When stochastic frontier production function model is employed to estimate the economic output effects of factor inputs, we need to first assess the appropriateness and specific form of stochastic frontier production function. Using likelihood ratio test and significance test, we find that the crossmultiplying term between capital stock and intermediate product input, the cross-multiplying term between technology level and capital stock, as well as the cross-multiplying term between technology level and intermediate product input, should be excluded, and final results are shown in Table 2 below. It can be seen that the coefficients of all independent variables of the model are significant at least at
10% significance level, with γ value as high as 0.9693 and significant at 1% level. This indicates that technology inefficiency generally exists, and that the error of frontier production function is primarily caused by technology inefficiency, which further demonstrates that the use of stochastic frontier production function is necessary and valid.
Relevant results of Table 2 provide possibilities for the analysis of factor inputs of 2015. Before specific analysis, this paper specifies capital stock as a discretionary variable given the universal existence of capital factor overcapacity in manufacturing sectors. This is intended to reduce capital factor overcapacity through factor structure matching. In order to follow the above analytical approach, this paper specifies the labor intensity coefficient of specific years as independent from the level of economic output, i.e. labor intensity coefficient obtained from the previous section can be used to calculate the employment that each sector is able to sustain based on optimized output level. Lastly, considering the fundamental supportive role of intermediate product inputs in manufacturing process, the intermediate product input intensity coefficients of specific years can be specified as independent from the level of economic output, so as to obtain the level of intermediate product input based on the optimized output level.
(2) Factor structure matching result of manufacturing sectors and analysis
Based on the desirable output values of manufacturing sectors, we calculate the reasonable levels of capital stock, labor and intermediate product input. Manufacturing industry as a whole requires a capital stock worth 14,708.56 billion yuan and intermediate product inputs worth 48,506.76 billion yuan, down
26.02% and up 0.74% compared with original values respectively, and may provide an employment extremely close to the original value of 2015. Comparatively speaking, manufacture of communication equipment, computers and other electronic equipment, manufacture of transport equipment, and manufacture of electrical machinery and equipment provide the most jobs, and use the most intermediate product inputs. This suggests that these three sectors play an extremely important role in promoting employment and supporting the development of other sectors.
In order to compare the increases and decreases of capital stock of various manufacturing sectors, this paper carries out an analysis by classifying sectors into those with absolute strong investment increase, weak absolute investment increase, relative investment increase, absolute investment reduction, strong relative investment reduction and weak relative investment reduction, with classification criteria shown in Table 3.
Based on Table 3 and Figure 3, we find that six sectors including manufacture of medicines, manufacture of transport equipment, manufacture of communication equipment, and computers and other electronic equipment are sectors of absolute strong investment increase; 11 sectors including manufacture of rubber, manufacture of raw chemical materials and chemical products, and manufacture of textile are sectors of absolute investment reduction; four sectors including manufacture of leather, fur, feather and related products, manufacture of textile wearing apparel, footwear and caps, and manufacture of non-metallic mineral products are sectors of strong relative investment reduction, while eight sectors including processing of food from agricultural products, manufacture of beverage, and recycling and disposal of waste are sectors of weak relative investment reduction.
In comparison between the size of capital stock and the size of output, we may notice a consistent trend in their changes, with the exception of some sectors. We notice that iron and steel, electrolytic aluminum, cement, coal chemicals, fan equipment, polycrystalline silicon and paper-making sectors are considered as sectors with serious overcapacity. In industry classification, they correspond to six sectors, including smelting and pressing of non-ferrous metals, smelting and pressing of ferrous metals, manufacture of raw chemical materials and chemical products, manufacture of paper and paper products, manufacture of non-metallic mineral products and manufacture of special purpose machinery (Dong et al., 2015). In the foregoing output structure optimization, the first four sectors all require a slowdown in output growth and a more substantial reduction in the size of capital stock; but such a reduction is merely an adjustment of the size of capital stock under the condition of accepting historical overcapacity and obsolete capacity. The following section will examine to what extent such an adjustment is able to resolve the problem of overcapacity.
4.3 Estimation and Reduction of Manufacturing Capital Factor Overcapacity
The Chinese government has attached great importance to addressing obsolete capacity in manufacturing industry, and achieved initial results. According to the NBS survey of 60,000 large and medium-size enterprises since 2014, almost all enterprises have capacity utilization rates below 80%. As China’s economy enters into the new normal, if slowing growth is not matched by a capital stock adjustment, the problem of overcapacity will persist and deteriorate. Hence, it is of great significance to assess the capital stock of China’s manufacturing industry.
This paper will employ input- oriented non- discretionary variable model with constant return to scale created by Cooper et al. (2004) to estimate the capital factor capacity utilization of various manufacturing sectors, including original values and optimized values of 2015. In order to create an efficiency frontier for each sector, relevant data of various economies is required. Considering data availability, this paper conducts an analysis of sector-specific data of 30 provincial regions and national
overall data with 31 DMU input-output data entries as samples.
(1) Estimation of capital factor overcapacity before and after manufacturing factor matching
Figure 4 reports the original and optimized values of capacity utilization of manufacturing sectors in 2015. The result shows that manufacturing industry’s original overall utilization is about 56.14%. Relatively, light industries and high-tech industries boast higher capacity utilization rates. For instance, medicine manufacturing, manufacture of textile wearing apparel and footwear and caps rank relatively high. However, the capacity utilizations of heavy industries are relatively low. For instance, recycling and disposal of waste, smelting and pressing of non-ferrous metals and processing of petroleum, coking, processing of nuclear fuel rank as the bottom three manufacturing sectors with the lowest capacity utilization rates. Overall rankings of capacity utilization rates of various sectors calculated in this paper are generally consistent with Han et al. (2011) and Dong et al. (2015).
Through optimization of manufacturing industry’s output structure and factor structure in 2015, we are able to greatly increase manufacturing industry’s utilization rate to an overall mean value of 72.04%. This value is still smaller than the desirable capacity utilization level often referenced by developed countries like the U. S. ( Zhong and Pan, 2014) by seven to ten percentage points. From a sectorspecific perspective, except for medicine manufacturing whose capacity utilization remains almost unchanged after optimization, capacity utilization rates have more or less increased for all other sectors. In particular, capacity utilization of textile industry may increase to 85.84%, which is the highest. This paper notices that the capacity utilization of recycling and disposal of waste is still as low as 55.12% after optimization, which is among the lowest. The reason is that in responding to the policy to develop “venous industry,” various localities fell into low-level repetitive construction and vicious competition. Developing “venous industry” is an inevitable choice for China, but issues related to the cross-regional transportation of waste resources should be addressed. “Venous industry” should be developed according to various factors such as population density, business density and cost of transportation.
(2) Reduction of manufacturing capital factor overcapacity
Manufacturing industry’s capacity utilization may increase after matching the factor structure of optimized desirable output merely by extracting historical information. However, overcapacity still exists. Here, capital stock of various manufacturing sectors is adjusted based on the non-linear relationship between capital stock and economic output in historical sample data. Adjusted capacity utilizations result should be equivalent to historical mean values. Capacity utilization rates of 2008-2010 in Figure 4 are 73.27%, which is rather close to the value after factor structure matching for 2015 (72.04%). In this sense, adjustment based on historical samples technically only approached the historical mean value. In an attempt to shore up slowing economy, local governments resorted to an investment spree. But this is only the first level for manufacturing industry to increase capacity utilization and reduce overcapacity. After capital stock is adjusted to historical mean value according to the level of desirable output, there is often a gap with the capacity utilization with domestic high-efficiency economy as efficiency frontier. Reducing this gap becomes the second level where China’s manufacturing industry may increase capacity utilization and reduce overcapacity.
On the basis of resolving overcapacity at the first level, we should focus on and resolve the following problem: The capacity utilization rates of recycling and disposal of waste, smelting and pressing of non-ferrous metals, processing of petroleum, coking, and processing of nuclear fuel are significantly smaller than desirable levels, and further adjust capital stock value based on their gaps to achieve the transition of capacity utilization from level 1 to level 2. This transition is a qualitative change, and will be much more difficult than achieving level 1. As noted by Coellie et al. (2002) and Dong et al. (2015), since various economies may have equal fixed inputs but different productivities (i.e. difference in technical efficiency), capacity utilization may be further decomposed into equipment utilization and technical efficiency (Coelli et al., 2002; Dong et al., 2015). Then, the most direct solution to capital factor overcapacity is to determine a reasonable size of firms based on output requirement
to avoid diseconomies of scale arising from excessively small or large scale, focusing on equipment utilization improvement. In addition to improving firms’ technical and managerial levels, as well as revealed technical efficiency, it is also important to reduce implicit technical inefficiencies by phasing out obsolete capacities and “bubble fixed assets” with exaggerated cost and value.
5. Concluding Remarks and Policy Implications
Reasonable industrial structure is the key to industrial behaviors and performance. In this sense, whether China is able to optimize its industrial structure according to its national conditions is vital to the success of “Made in China 2025.” Using two-digit manufacturing sectors, this paper systematically optimizes China’s manufacturing industrial structure. Our findings suggest that (1) manufacturing output structure optimization may reduce energy intensity and carbon intensity by 18.08% and 17.42% respectively; (2) to reduce factor mismatch, input factors need to be matched after manufacturing output structure improvement. The level of capital stock, in particular, requires a 26.02% reduction; (3) estimation result of capital factor capacity utilization further reveals that China’s manufacturing capacity utilization in 2015 was far below the average level in the mid-and late stage of the 11th Five-Year Plan period (2008-2010). After input factor matching, capacity utilization may rise to the latter level.
Based on this paper’s findings, we may arrive at the following policy implications: First, “Made in China 2025” strategy should not be intended for all sectors indiscriminately. Instead, there should be priorities for the development of specific sectors and retreat of some others. We suggest giving priority to developing nine sectors, including manufacture of medicines and manufacture of special purpose machinery, to speed up economic growth; properly controlling the growth rates of manufacture of beverage, manufacture of cultural, educational, fine arts, sports and entertainment goods, as well as other manufacturing (these sectors should outpace manufacturing industry’s benchmark growth rate); growth rates of other manufacturing such as processing of food from agricultural products and manufacture of foods should stay below manufacturing industry’s benchmark growth rate and avoid excessive growth.
Second, the heterogeneity of production factors requires the government and market to play different roles in synergy. The government should eliminate institutional labor market segregation, reduce institutional privileges related to household registration ( hukou) and quota, and thus lower the cost of labor migration and promote the free flow of labor. However, potential unemployment arising from excessive increase of profit-seeking capital component should be avoided. According to the needs of output structure optimization, the central government should identify a reasonable level of capital stock for manufacturing sectors, avoid disproportionate investment in overall sector planning, and regulate and restrain local government investment to curb excessive investments. The government should withdraw from its role as an investment entity, and refrain from intervening in capital factor allocation. Instead, it should regulate and guide the market, and ensure market-based investment activities.