1
MASTER THESIS
MSc. International Economics and Business
CHINESE LABOUR PRODUCTIVITY,
A STRUCTURAL DECOMPOSITION ANALYSIS
FROM 1995-2008
2
Abstract
In this thesis Chinese labour productivity growth from 1995 to 2008 is decomposed in six partial factors (both supply and demand): (1) changes in value added coefficients , (2) labour inputs , (3) shares of sectoral demands that are fulfilled domestically , (4) input mix , (5) the intra-sectoral shares of final demand and, (6) the inter-sectoral mix of final demand . The results show structural change has taken place in the Chinese economy during the period 1995-2008. Furthermore, partial factor E has had the biggest contribution in labour productivity growth for all sectors. This has been partly caused by the more flexible Chinese labour market. Finally, no support has been found for a positive influence of export on the labour productivity
3
Table of Contents
1. Introduction 5 2. Review of Literature 7 3. Methodology 15 4. Data 22 5. Results 27 6. Conclusion 39 7. Discussion 42 8. References 43List of Tables and Figures
Figure 1 14
Growth rates of real exports, from 1995-2008 (constant prices in US$) (%)
Figure 2 27
Annual employment share per aggregated sector, for 1995-2008 (%)
Table 1 22
WIOD industries for China used in this research
Table 2 23
Subdivision of the industry sector
Table 3 24
Subdivision of the services sector
Table 4 29
4
Table 5 30
Yearly labour productivity levels, annual labour productivity growth rates per period for the total economy (%), value added per year in million US$ and the annual growth rates of value added per period (%).
Table 6 30
Labour productivity levels and labour productivity growth per aggregate sector, per year and per period (%)
Table 7 31
Factor contribution of total economy and per aggregate sector (%)
Table 8 32
Factor contribution per subsector of the industry sector, textiles and food (1I), wood and paper (2I), petroleum, chemicals and metals (I3), machinery and electronics (I4) and manufacturing and construction (I5) (%)
Table 9 34
Factor contribution per subsector of the services sector, trade (1S), transport (2S), communication, finance and real estate (3S), government services (4S), other services (5S) (%)
Table 10 35
Factor contribution per subsector of the industry sector, second division: market services (MS) and government services (GS) (%)
Table 11 36
5
1. Introduction
China has been able to increase its per capita income substantially and it has developed itself to one of world’s biggest economic players over the last decades. Today, the Chinese economy is world’s second largest economy by producing approximately 11.3 percent of global GDP in 2013 (World Bank, 2014). It could achieve this by attaining a fast growth in GDP and labour productivity over the past decades. Chinese labour productivity has grown at a spectacular rate of 8.3 percent during 1995-2006 and 9.6 percent during 2007-2012 (World Bank, 2014). Such a fast aggregate growth in productivity growth could be accrued either by a productivity gain in individual sectors of the economy, or via a positive structural change, where resources move from low productive to high productive sectors. The role of structural change – reallocation of labour from low to high productivity sectors – in accelerating aggregate growth has been given significant importance in the literature on economic growth (Lewis, 1954; Kuznets, 1966; Chenery et al. 1986; Fan et al., 2003; McMillan, Rodrik, 2011).
After the economic reforms in 1978 and the establishment of special economic zones (SEZs), extensive inter-regional migration started in Chinese economy (Lin, 2011). Workers moved from central provinces that were focusing primarily on agricultural activities to coastal provinces, focusing on other modern economic activities, (Rodrique et al, 2013). This led to a wholesale structural shift across the country from low-productivity sectors towards those with higher productivity (Fan et al., 2003). In addition to this large degree of structural change, there is observed productivity gain across most individual sectors through the early to mid-1990s, which was achieved by several federal policies and increasing (foreign) investments. These developments have led to diverse development patterns across industries (Yang and Lahr, 2010).
6 et al (2003) divide the Chinese economy into four different industries, where Yang and Lahr (2008, 2010) identify five to eleven different sectors. An exception is de Vries et al (2012), who make a comparative study of the BRIC countries. They analyse 35 different industries, using growth accounting. However, de Vries et al. only include the supply side of the story, and therefore they are neglecting the inter-industry linkages. This thesis focuses on China specifically and takes a more disaggregated approach, exploiting a large database of detailed sector data to answer the following research question:
Which sectors are driving China’s aggregate economic growth during 1995- 2008 and which factors are driving the growth of these sectors?
The focus lies on the period 1995-2008 for two reasons. First, this time frame has seen many dynamics in both the Chinese as well as the global economy – such as liberalization of trade, declining communication and transport costs and a new Chinese labour rule – on which we will elabourate in a later section (Yang and Lahr, 2010; Lin, 2011). Secondly, this is the time frame for which the largest sample of disaggregated data is available. Our analysis provides more recent evidence on Chinese growth dynamics, as most researches currently available are only until the late 1990s (Fan et al, 2003; Yang and Lahr; 2008) or the early 2000s (Kuijs and Wang, 2005; Yang and Lahr, 2010). Using an input-output structural decomposition analysis, labour productivity growth is decomposed into six partial factors (both supply and demand side factors) – (1) changes in value-added coefficients, (2) labour inputs, (3) shares of sectoral demands that are fulfilled domestically, (4) input mix and (5) the intra-sectoral shares and (6) intersectoral mix of final demand. This helps us identify what factors contribute most to China’s aggregated productivity growth over the years. For this purpose, we use the World Input Output Database (WIOD), which provide a coherent time series data on input output tables and socio-economic accounts for the period 1995-2009.1 . However, as the global financial crisis started in 2008, we opted to choose both 2007 and 2008 as end years of our analysis to make sure it is not influenced by the crisis.
The thesis is organized as follows: in chapter 2 a literature review is presented
1 For many variables WIOD data is available until 2011, however, since employment data is available
7 in which the hypotheses are also introduced. In chapter 3 and 4 the methodology and the data description are presented. In section 5 the results of the decomposition analysis are provided. Chapter 6 and 7, conclude the thesis.
2. Review of Literature
In this section we provide a review of literature, after which the hypotheses are presented. First of all, the concept of structural change is introduced. Then the changes in the global and Chinese economy are discussed. Subsequently, Chinese growth is discussed both from the growth accounting as well as the structural change perspective. Lastly, the hypotheses are presented.
In 1954, Arthur Lewis formulated the concept of development as “a transition from traditional to modern ways of production and economic behaviour”. During this transition period the supply of unskilled labour is elastic; profits, savings and investments are increasing; industry develops more rapidly than agriculture; and finally the pattern of international trade gradually transforms as the comparative advantage of the country changes (Lewis, 1954). In 1966, Simon Kuznets observed that all countries that undergo economic development show very high rates of output growth and similar shifts in resource allocation. Chenery et al. (1979) extended the ideas of Lewis (1954) and Kuznets (1966) that were designed to provide an empirical basis for models of development (Chenery and Watanabe, 1958; Chenery 1960, 1964; Chenery and Taylor 1968; Chenery and Syrquin, 1975). All researches support the fact that large economic changes are occurring in basically all economic factors during transition. Firstly, the production capacity increases i.e. accumulation of capital and skills. Secondly, there is a transformation of resource use i.e. demand, production, trade and factor use. Lastly, socioeconomic processes such as urbanization, the distribution of income and demographic transition are observed (Chenery and Watanabe, 1958; Chenery, 1960, 1964; Chenery and Taylor 1968; Chenery and Syrquin, 1975).
8 theory however, primary attention is given to the factor level i.e. the allocation of labour and capital (Chenery, 1960). Commodities are significant to a lesser extent and only when they show different production or demand functions. The transformation of factor use can be divided into three components: (1) a change in overall factor proportions through the accumulation of physical capital and skills; (2) reallocation of these factors among productive sectors; (3) increases in total factor productivity per sector (Chenery, 1960).
When a country is developing, specific patterns can be identified. At industry level, technological change fosters differential patterns of sectoral productivity growth. Additionally, changes in domestic demand and international trade patterns drive a process of structural transformation, in which intermediate inputs, labour and capital are constantly being shifted between companies, sectors and countries (Kuznets, 1966; Chenery et al., 1986; Harberger, 1998; Hsieh and Klenow; 2009). The nature and speed at which this structural change, or the movement of resources across sectors, takes place is one of the key determinants to the success of a country (McMillan and Rodrik, 2011). Based on this, new structural economists argue for production structures to be the new starting point for comparative economic analysis and the design of appropriate policies (Lin, 2011).
9 was quite rapid; while about 74 percent of Chinese population lived in rural areas in the 1990s, it declined to about 27 percent in 2009 (World Bank, 2014). China changed from a closed market to a market that is open and eager to trade. One of the major steps in this process was China’s accession the WTO in 2001. All these may have substantial impact on China’s transformation and growth. Chinese annual economic growth during 1995-2008 was 11.1 percent (Worldbank, 2014), and its annual labour productivity growth (output per hours worked) during this period was 10.2 percent (World Input Output Database, 2014). Moreover, China’s employment share in agriculture declined from 51 percent in 1997 to 40 percent in 2008 (de Vries et al., 2012).
10 China’s economic growth. Lin (2011) presents a very comprehensive research on the reasons for economic growth over the past 30 years. The first reason he provides is China’s adoption of a dual-track approach – in which the government controls key sectors of the economy, while allowing private enterprises limited control over other sectors. Through this approach China was able to achieve stability and dynamic transformation simultaneously (Lin, 2011). Secondly, China was a latecomer, developed according to its comparative advantage and tapped into the potential advantage of backwardness. The main difference between Lin (2011) and the other authors discussed above (e.g. Kim and Lau, 1996; Nehru et al, 1997; Maddison, 1998; Holz, 2006a) is that he links his story on China’s economic growth to the concept of structural change. Although, he does not provide any analysis regarding structural change he clearly highlights the importance of the concept. In conclusion, the larger part of current literature argues increasing capital stock in combination with better use of resources and a substantial growth in total factor productivity has driven Chinese economic growth. The main critique on these researches is that many use growth accounting to measure the contribution of different factors to economic growth, based on Solow (1957). By using this approach, inter-industry linkages are largely ignored which is highly disadvantageous to such an analysis, as only the supply side of the story is taken into account (Yang and Lahr, 2010). This could lead to a distorted picture as the demand side also influences the economy greatly; think of more liberal trade regulations for instance. These new regulations will trigger a growing demand for certain products or services, which will lead to an increase in output. Therefore, in order to provide a complete picture, both the supply as well as the demand side should be taken into account.
11 (2003) largely ignore the interdependence of China’s industries by considering only the supply side of the story. Bosworth and Collins (2008) make use of growth accounting as well and divide the industry into three different sectors – primary, secondary and tertiary. They conclude that the secondary sector (i.e. industry) has accounted for 60 percent of total productivity growth, which is in line with the general literature regarding the secondary industry (Bosworth and Collins, 2008). Bosworth and Collins argue that 1.2 percent of total factor productivity was caused by structural change. Again we see an aggregate analysis of only three economic sectors, which masks industry heterogeneity. Besides, Bosworth and Collins (2008) also only look at the supply side of the story, and therefore neglect the inter-industry linkages.
Contradictory to the researches discussed previously (e.g. Kim and Lau, 1996; Nehru et al, 1997; Maddison, 1998; Fan et al., 2003; Holz, 2006a; Bosworth and Collins, 2008) there do exist some studies that have used detailed data at sector level. De Vries et al. (2012), perform an industry analysis for China by examining 35 different industries using a decomposition technique suggested by Fabricant (1942). They find a declining contribution to productivity growth for the agricultural sector between the years 1987 and 2008 in China, as found by Yang and Lahr (2010) as well. In addition many manufacturing sectors witnessed an increase in their contribution, with electrical and optical equipment being among the largest contributing sector (De Vries et al., 2012). In services on the other hand, though overall employment share is increasing, it is highly concentrated in below average productive sectors such as retail trade and other community and personal services, thus leaving a slower structural change effect (de Vries et al., 2012). Nevertheless, the analysis of de Vries et al. (2012) also has the flaw to only include the supply side of the story, as was the case with other researches discussed before (e.g. Kim and Lau, 1996; Nehru et al, 1997; Maddison, 1998; Fan et al., 2003; Holz, 2006a; Bosworth and Collins, 2008).
12 by fostering substantial economic structural transformation (Fan et al, 2003; Bosworth and Collins, 2008; Yang and Lahr, 2010; De Vries et al, 2012).
In conclusion, not much research has been conducted on structural change in China, especially for the post 1995 period. The lack of detailed sector data complicates a thorough assessment of the importance of structural change for productivity growth in China specifically. Often only aggregated data is available, with the industry divided into three main sectors (1) agriculture, (2) manufacturing and (3) services. Such an aggregate picture masks substantial heterogeneity at industry level.
Given the significant growth surge in Chinese output and productivity and the on-going shift from agriculture to manufacturing and further a service driven economy, we postulate the following hypothesis:
Hypothesis I: Structural change significantly contributes to the increase in labour productivity growth and hence to China’s aggregate growth pattern for the period 1995-2008.
In addition to determining whether structural change has a positive contribution to the Chinese labour productivity, Yang and Lahr (2010) provide a method to identify the drivers for this structural change. In their research they identify six partial factors (1) changes in value added coefficients, (2) labour inputs, (3) shares of sectoral demands that are fulfilled domestically, (4) input mix, (5) the intra-sectoral shares of final demand and, (6) the inter-sectoral mix of final demand. By performing a structural decomposition analysis based on input-output tables, total labour productivity is decomposed into these six partial factors. Based on their analysis, labour inputs seem to have had the largest contribution to total labour productivity, followed by the intra-sectoral shares of final demand and the inter-sectoral mix of final demand. By performing this analysis one could have a better understanding of that factors that enable sustained GDP growth in China (Yang and Lahr, 2010).
13 during their research period, from 108.8 percent in 1987-1997 to 84.9 percent in 1997-2002 and 82.9 percent in 2002-2005. Hence, returns to capital deepening appear to have attained decreasing returns to scale in China (Yang and Lahr, 2010). In addition, we need to take into account that China started to trade in the global competitive world market, where the returns to scale are lower due to the heavy competition (Yang and Lahr, 2010). Besides, they highlight the importance of the reforms that have taken place in the employment market in 1986. In this year, the Chinese government imposed a regulation that state owned enterprises could hire personnel via contracts – which meant that now state-owned enterprises could not only hire but also fire labourers. In 1995, this labour contract system was adopted for the entire market. However, it took until 1998 for the Chinese government to allow employment to face the full force of the market economy (Yang and Lahr, 2010). It seems likely that the remaining workforce is stimulated to work hard and efficiently in order to retain their job. This ‘pressure’ has led to an significant quality increase in the Chinese workforce after 1995. To give an indication, 751,347 students were enrolled for level third education in 1995, where 1,337,455 were in 2008 (China Labour Statistical Yearbook, 2012). As a result of this quality improvement of the labour force, labour-saving effects should have further increased. A limitation of this research is that Yang and Lahr (2010) have only little data on the services sector. Four different service industries are distinguished: “Finance and Insurance”, “Real Estate”, “Wholesale and Retail Trade” and “Catering Trade”. The other data regarding the service sector is grouped in “Other Services”. These 5 ‘industries’, are divided among two sectors for their final analysis, which does not provide a very strong picture.
Given the findings in previous literature (Yang and Lahr, 2010; De Vries et. al., 2012), we postulate the following hypothesis:
Hypothesis II: the increased flexibility in the labour market, due to the liberal market reforms, has a positive influence on productivity, by reducing the labour use per unit of output.
14 0% 20% 40% 60% 80% 100% 120% 140% 70 63 AtB M O C 21 t2 2 20 H E L 15 t1 6 19 J 17 t1 8 26 60 25 N F 62 27 t2 8 64 36 t3 7 23 30 t3 3 61 29 24 34 t3 5 71 t7 4 51 52 Annual growth rates of real exports (constant prices in US$)
to the world. The demand for China’s products and services rose tremendously from 222,759 million US$ (deflated to 2009) in 1995 to 1,593,709 million US$ (deflated to 2009) in 2008; the share of exports per total output rose from 8.9 percent in 1995, to 11.4 percent in 2008 (World Input Output Database, 2014). Therefore, it is highly important to take the demand side into account as well when analysing labour productivity.
Figure 1: Growth rates of real exports, from 1995-2008 (constant prices in US$) (%)
source: World Input Output Database, (Timmer, 2012)
Figure 1 provides the trend in exports in Chinese industries during 1995-2008, which clearly indicates a growth in exports for nearly all industries. Sector 51 and 52, both services sectors, experience the largest export growth in the period 1995 to 2008. Exports that amounted 0.008 million US$ for sector 51, and 0.001 million US$ for sector 52 in 1995; have skyrocketed in the subsequent years, leading to a growth rate of approximately 1,600 percent. The third largest sector in terms of export growth rates is 71t74 (Renting of M&Eq and Other Business Activities) – which is also a service sector – with a growth rate of 359 percent.
15 (2010), we hypothesize that:
Hypothesis III: The export intensive industries in Chinese economy, which are highly exposed to international competition, have contributed significantly to China’s aggregate productivity growth.
3. Methodology
To see whether this labour productivity growth can be attributed to growth within industries, or whether this growth is caused by one of the partial factors, a structural decomposition analysis is made. In literature, often growth accounting is used to assess drivers for Chinese economic growth (Kim and Lau, 1996; Nehru et al, 1997; Maddison, 1998; Holz, 2006a; Bosworth and Collins, 2008; Lin, 2011; De Vries et al., 2012). However, when using this approach, inter-industry linkages are neglected as the demand side of the story is not taken into account (Yang and Lahr, 2010). For that reason we make use of a structural decomposition analysis based on input-output tables.
We use a decomposition method provided by Yang and Lahr (2010), based on Jacob’ (2003), where labour productivity is decomposed into six partial factors: (1) changes in value added coefficients , (2) labour inputs , (3) shares of sectoral demands that are fulfilled domestically , (4) input mix , (5) the intra-sectoral shares of final demand and, (6) the inter-sectoral mix of final demand . The aim of this research is to examine to what extent structural change contributes to the increase in Chinese labour productivity and what partial factors have had the largest impact on this increase.
16 where subscript 1 indicates the final year of the period being either 2001, 2007 or 2008.
This type of decomposition analysis is only suitable for aggregate industries, because then it can show the shifts in the partial factors: (3) shares of sectoral demands that are fulfilled domestically , (4) input mix , (5) the intra-sectoral shares of final demand and, (6) the inter-sectoral mix of final demand. One cannot perform a decomposition analysis on a single industry, as there will be no influence of shifts in these abovementioned partial factors for one industry.
We will make multiple decomposition analyses in this research. First we will distinguish three different industries: agriculture, industry and services, to see whether structural change can be observed at a very general level. We applied the division as used in table 1 (see “Data”). Subsequently, we will assess the industry and services sector in more detail by making subdivisions of these industries, represented by table 2 and 3 (see “Data”).
When working with structural change decomposition, one also has to use the other polar decomposition with reverse weights (Dietzenbacher and Los, 1998). This is because the structural change decomposition is not unique. As mentioned by Dietzenbacher and Los (1998), taking the average of these two polar decompositions gives a value that is very close to the average of all the possible decomposition forms. The other polar decomposition with reversed weights is represented by equations 1.1r to 1.6r.
Consider,
n represents the number of industries;
v vector of value added (n×1 vector) (specify real or nominal value added); e vector of labour inputs (n×1 vector);
λ labour productivity (where λ = vi/ei) (n×1 vector);
A matrix with input coefficient (n×n matrix), with typical element of aij
denoting the input of product i per unit of output in industry j;
17 of the corresponding cell in the final demand matrix to its respective column
sum;
y aggregate final demand for each of k categories (household consumption, government consumption, gross fixed capital formation,
inventory stock) (k×1 vector);
diagonal matrix with elements ei as the labour input per unit of output in
industry i in the diagonal and “0” elsewhere (n×n matrix);
diagonal matrix with elements vi as the value-added per unit of output in
industry i in the diagonal and “0” elsewhere (n×n matrix); and
diagonal matrix with elements ρi as the domestic supply ratio (ratio of the
total output minus export to total supply, that is, total output minus export and then plus import) in industry i in the diagonal and “0” elsewhere (n×n
matrix).
This research examines: which sectors are driving China’s aggregate economic growth during 1995-2008 and which factors are driving the growth of these sectors?
To answer this question we consider six different partial factors. Given the significance of structural change for economic growth, and the fact that China has been growing very rapidly, it is imperative to understand what factors are driving the aggregate economic growth in China and whether the resource movements across different sectors are growth enhancing. The six partial factors we distinguish in this research are:
1) changes in value added coefficients , diagonal matrix with value added per unit of output
2) labour inputs , i.e. diagonal matrix of labour inputs per unit of output 3) shares of sectoral demands that are fulfilled domestically , ratio of total
output minus export, plus import
4) input mix , the matrix of direct input coefficients derived based on both domestic and imported intermediate inputs employed in the production process
18 6) the inter-sectoral mix of final demand , vector of aggregate demand by final
use (excluding exports)
Bex is the vector of the normalized export coefficients. Lastly, ex is the scalar of the aggregate value of exports.
The value added vector can be written as follows,
This leads to the following equation that forms the basis for the main equations in this research,
The main equations are as follows,
where indices are time indicators,
19 term is added. L0 implies that and for time period 0 are taken into account; L1 implies that and for time period 1 are taken into account; L01 implies that for time period 0 and for time period 1 are taken into account and finally, L10 implies that for time period 1 and for time period 0 are considered.
We can decompose the change of value added as:
Similarly we derive the decomposition for the change of labour inputs. By combining these two decompositions, we have the decomposition equation for the change of labour productivity:
Labour productivity is calculated as the ratio of value added to employment, i.e.
20 domestic supply ratio. Equation 1.4 gives the effects of changes in the inter-industry structure that is due to technical change. In addition, equation 1.5 assesses the intra-sectoral shares of final demand and. Lastly, equation 1.6 shows the changes in inter-sectoral mix of final demand. The six equations are shown below,
(1.1)
changes in value added per unit of gross output
(1.2)
changes in labour inputs
(1.3)
changes in shares of sectoral demands fulfilled domestically
(1.4)
changes in input mix
(1.5)
changes in intrasectoral shares of final demand
(1.6)
changes in intersectoral shares of final demand
The formulas 1.1 until 1.6 result in vectors (nx1), which means there will be an outcome per aggregate industry (for the exact number of industries per decomposition please consult table 2 to 4). In addition, when taking the summations of the vectors in the nominator and the dominator, and by subsequently dividing these by each other, the result will be a scalar of the total value.
21 where r indicates the polar decomposition with reversed weights
(1.1r)
changes in value added per unit of gross output
(1.2r)
changes in labour inputs
(1.3r)
changes in shares of sectoral demands fulfilled domestically
(1.4r)
changes in input mix
(1.5r)
changes in intrasectoral shares of final demand
(1.6r)
changes in intersectoral shares of final demand
By taking the natural logarithm the percentage contribution for each factor is achieved.
22 2002-2008.
4. Data
In order to conduct this research data is needed on (1) value added per sector, (2) labour hours worked per person engaged, (3) input values from sector i to j in matrix form, (4) (sector specific) final demand, (5) imports and exports. All required data can be found in the World Input Output Database (WIOD) (Timmer, 2012).
World-Input-Output-Database (WIOD) is a publicly available database that provides time-series of input-output tables for 40 countries worldwide for the period 1995-2011.2 In addition, labour and capital inputs and environmental indicators at the industry level are available. The WIOD complies with national accounts and international trade statistics. The database is constructed in monetary terms (millions US dollars). The combination of the National-Input-Output-Table of China with the data on labour inputs provides the relevant data required for this research. Even though WIOD consists of 35 industries in general, in the case of China two sub-sectors (ISIC 50 and P) are not included in this research due to lack of data. Therefore the current analysis is confined to 33 sectors listed in Table 1; the first decomposition analysis is made on a three sector level, similar to this classification.
Table 1: WIOD industries for China used in this thesis
Number ISIC rev. 3 Description 3-sector
1 AtB Agriculture, Hunting, Forestry and Fishing Agriculture
2 C Mining and Quarrying Agriculture
3 15t16 Food, Beverages and Tobacco Industry
4 17t18 Textiles and Textile Products Industry
5 19 Leather, Leather and Footwear Industry
6 20 Wood and Products of Wood and Cork Industry 7 21t22 Pulp, Paper, Paper , Printing and Publishing Industry 8 23 Coke, Refined Petroleum and Nuclear Fuel Industry
9 24 Chemicals and Chemical Products Industry
10 25 Rubber and Plastics Industry
11 26 Other Non-Metallic Mineral Industry
12 27t28 Basic Metals and Fabricated Metal Industry
13 29 Machinery, Nec Industry
14 30t33 Electrical and Optical Equipment Industry
23
15 34t35 Transport Equipment Industry
16 36t37 Manufacturing, Nec; Recycling Industry 17 E Electricity, Gas and Water Supply Industry
18 F Construction Industry
19 51
Wholesale Trade and Commission Trade, Except of Motor Vehicles and
Motorcycles Services
20 52
Retail Trade, Except of Motor Vehicles and Motorcycles; Repair of
Household Goods Services
21 H Hotels and Restaurants Services
22 60 Inland Transport Services
23 61 Water Transport Services
24 62 Air Transport Services
25 63
Other Supporting and Auxiliary Transport Activities; Activities of Travel
Agencies Services
26 64 Post and Telecommunications Services
27 J Financial Intermediation Services
28 70 Real Estate Activities Services
29 71t74 Renting of M&Eq and Other Business Activities Services 30 L Public Admin and Defence; Compulsory Social Security Services
31 M Education Services
32 N Health and Social Work Services
33 O Other Community, Social and Personal Services Services source: NIOT of China, World Input Output Database (2014)
Then we focus on the industry sector specifically. By grouping the 16 individual industry sectors into 5 aggregate sectors, we form the dataset for the second decomposition analysis (table 2).
Table 2: Subdivision of the industry sector
Index ISIC Sector Industries
1I 15t16, 17t18, 19 Textile and food products Food, Beverages and Tobacco; Textiles and Textile Products; Leather, Leather and Footwear
2I 20, 21t22 Wood and paper products Wood and Products of Wood and Cork; Pulp, Paper, Printing and Publishing
3I 23, 24, 25, 26, 27t28
Petroleum, chemicals and metals Coke, Refined Petroleum and Nuclear Fuel; Chemicals and Chemical Products; Rubber and Plastics; Other Non-Metallic Mineral; Basic Metals and Fabricated Metal
4I 29, 30t33 Machinery and electronics Machinery, Nec; Electrical and Optical Equipment 5I 34t35, 36t37, E, F Manufacturing and construction Transport Equipment; Manufacturing, Nec; Recycling;
Electricity, Gas and Water Supply; Construction source: World Input Output Database (2014)
24 services.
Table 3: Subdivision of the services sector
Index ISIC Sector Industries
1S 51,52 Trade Wholesale Trade and Commission Trade, Except of Motor Vehicles and Motorcycles; Retail Trade, Except of Motor Vehicles and Motorcycles;
Repair of Household Goods
2S 60, 61, 62, 62 Transport Inland Transport; Water Transport; Air Transport; Other Supporting and Auxiliary Transport Activities; Activities of Travel Agencies
3S 64, J, 70 Communication, Finance & Real Estate
Post and Telecommunications; Financial Intermediation; Real Estate Activities
4S 71t74, L, M, N Government services Renting of M&Eq and Other Business Activities; Public Admin and Defence; Compulsory Social Security; Education;
Health and Social Work
5S O Other services Hotels and Restaurants; Other Community, Social and Personal Services
MS 51, 52, 60, 61, 62, 63, 64, J, 70, 71t74,
L, M, N, O
Market services Wholesale Trade and Commission Trade, Except of Motor Vehicles and Motorcycles; Retail Trade, Except of Motor Vehicles and Motorcycles; Hotels and Restaurants; Repair of Household Goods; Inland Transport; Water Transport; Air Transport; Other Supporting and Auxiliary Transport Activities; Activities of Travel Agencies; Post and Telecommunications; Financial Intermediation; Real Estate Activities
GS 71t74, L, M, N, O Government services Renting of M&Eq and Other Business Activities; Public Admin and Defence; Compulsory Social Security; Education; Health and Social work; Other Community, Social and Personal services
source: World Input Output Database (2014)
For the decomposition analyses we perform in this thesis, data is needed on (1) value added per sector, (2) labour hours worked per person engaged, (3) input values from sector i to j in matrix form, (4) (sector specific) final demand, (5) imports and exports:
25 matrix , by dividing value added per unit of output.
Labour hours worked per person engaged
This information is derived from the Socio-Economic and Environmental Accounts (SEA) of the WIOD. In the SEA of China this variable is referred to as “H_EMP Total hours worked by persons engaged (millions)”. We derive at the labour inputs coefficients by dividing the labour inputs found in the SEA of China, by the total output found in the National Input Output Table (NIOT) of China.
Input values from sector i to j in matrix form
This data is available from the National Input Output Table (NIOT) for China. It is compiled of National supply and use tables at current and previous year prices (35 industries by 59 products) and National Input-Output tables in current prices (35 industries by 35 industries). The column indicates the values of all intermediate inputs used in production. Input values are presented in million of US$ in current prices. With this information we compile, input mix : the matrix of direct input coefficients derived based on both domestic and imported intermediate inputs employed in the production process.
(Sector specific) final demand
This data is available from the NIOT for China. Final use includes domestic use (private or government consumption and investment) and exports. Note, exports is considered separately in the calculations. The final element in each row indicates the total use of each product. Final demand is measured in million of US$ in current prices. We use this information to calculate the intra-sectoral shares of final demand - which is an array of the column normalized final demand coefficients that are both domestically produced and imported - and the inter-sectoral mix of final demand , - which is a vector of aggregate demand by final use (excluding exports).
Imports and exports
26 coefficients and ex – the scalar of the aggregate value of exports.
The IO-tables from the WIOD are all stated in current prices. In order to be able to compare the productivity levels over the years, it is necessary to correct for inflation. We divide current prices by the output price deflator with 2009 as a base this way the imbalance is largely offset. The GDP deflators used to compute real dollar values of output, value added, private and government consumption, private and government investment, exports and imports, are derived from the Worldbank (worldbank.org). It may not be a completely flawless method to control for economy wide price changes, as not all industries are subject to the same inflation rate. When inflation is not corrected properly, productivity growth will be overstated due to the fact that inflation costs are not taken into account.
27 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 19951996199719981999200020012002200320042005200620072008 Per ce n tage Year
Employment share 1995-2008 (hours worked by persons
engaged) (%)
Agriculture Industry Services
5. Results
In this section we present the results for the three different decomposition analyses we made and supporting information.
It has been established in the literature that the level and growth rate of labour productivity in agriculture is generally lower than in the rest of the economy, indicating differences in the nature of the production function, investment opportunities and the degree of technical development (Syrquin et al., 1984; Crafts, 1984). Therefore shrinking agriculture would indicate a faster decline in the contribution of this sector to output and productivity growth of the aggregate economy. Figure 2 shows the annual employment share per aggregated sector (respectively agriculture, industry and services). For China, the decline in the share of agriculture in aggregate GDP and employment has been reflected as an increase in the share of other sectors of the economy. The employment share of industry has increased from 26 percent in 1995 to 32 percent in 2008, while that of services increased from 30 to 35 percent (figure 2). In 2008, we observe that all three aggregate sectors are extremely close to each other in terms the percentage of hours worked by persons engaged (figure 2).
Figure 2: Annual employment share per aggregated sector, for 1995-2008 (%)
28 Figure 2 clearly shows that the employment has moved quite heavily away from the agricultural sector. Since 1995, workers are more and more moving out of the agricultural sector, as we expected from earlier literature (e.g. Fan et al. 2003; Yang and Lahr, 2010). However, what we expected to see was that the aggregated industry sector benefitted from this shift of workers before services did. However, when looking at the shares of employment per sector presented in table 4, it shows workers are moving from agriculture into services from 1995-2001. From 2002-2008, the workers are moving from agriculture into industry. This is not in line with the classical ideas on structural change that workers should first move into industry before moving into services (Chenery, 1960).
29
Table 4: Annual labour productivity growth rates per period (%)
# Sector Industry 1995-2001 # Sector Industry 2002-2007 # Sector Industry 2002-2008
1 S 61 33% 1 S 63 34% 1 S 63 31% 2 S 64 21% 2 S 70 18% 2 S O 17% 3 S 71t74 18% 3 S O 16% 3 S 70 17% 4 S L 15% 4 S 71t74 16% 4 S 51 17% 5 I C 15% 5 S J 16% 5 S 52 17% 6 S N 15% 6 S 51 15% 6 S J 17% 7 I 34t35 14% 7 S 52 15% 7 S 71t74 17% 8 I 23 14% 8 I 27t28 15% 8 I 36t37 16% 9 S M 13% 9 I 36t37 14% 9 I F 15% 10 I 29 13% 10 I F 14% 10 I 27t28 14% 11 I 24 13% 11 A AtB 12% 11 A AtB 14% 12 I 36t37 12% 12 I C 12% 12 S N 13% 13 S 62 10% 13 S N 12% 13 I C 13% 14 S O 10% 14 S M 11% 14 S M 13% 15 I E 9% 15 S L 11% 15 I 26 13% 16 I 26 9% 16 I 26 10% 16 S L 13% 17 I 27t28 9% 17 S 64 9% 17 S 64 12% 18 I 30t33 9% 18 I E 9% 18 I E 11% 19 S 60 8% 19 S H 8% 19 I 17t18 10% 20 S H 7% 20 I 17t18 8% 20 S H 10% 21 S 52 6% 21 I 19 8% 21 I 19 10% 22 S 51 6% 22 I 15t16 8% 22 I 15t16 10% 23 I 15t16 5% 23 S 60 6% 23 I 25 8% 24 I 17t18 5% 24 I 34t35 5% 24 S 60 7% 25 I 21t22 4% 25 S 61 5% 25 I 20 7% 26 S 70 4% 26 I 25 5% 26 S 61 7% 27 I 20 3% 27 I 20 5% 27 I 34t35 7% 28 A AtB 2% 28 I 30t33 4% 28 I 29 6% 29 I 25 2% 29 I 29 4% 29 I 30t33 6% 30 S J 2% 30 I 24 4% 30 I 24 5% 31 I F 2% 31 I 23 0% 31 I 23 3% 32 S 63 -2% 32 I 21t22 0% 32 I 21t22 3% 33 I 19 -4% 33 S 62 -6% 33 S 62 -4%
source: World Input Output Database (WIOD), 2014
30
Table 5: Yearly labour productivity levels, annual labour productivity growth rates per period for the total economy (%), value added per year in million US$ and the annual growth rates of value added per period (%).
Year Labour Productivity Value Added
1995 0.79 965447
2001 1.18 1581613
2002 1.22 1709337
2007 2.41 3590836
2008 2.96 4556121
Period Growth Rate Growth
1995-2001 6.7% 8.2%
2002-2007 13.6% 14.8% 2002-2008 14.8% 16.3% 1995-2008 10.2% 11.9% source: World Input Output Database (WIOD), 2014
Table 6: Labour productivity levels and labour productivity growth per aggregate sector, per year and per period (%)
Agriculture Industry Services
Year Labour Productivity Labour Productivity Labour Productivity
1995 0.44 1.31 0.87
2001 0.53 1.91 1.40
2002 0.53 2.06 1.43
2007 1.08 3.31 2.89
2008 1.36 4.00 3.53
Period Growth Rate Growth Rate Growth Rate
1995-2001 3.4% 6.3% 7.9%
2002-2007 14.3% 9.5% 14.1%
2002-2008 15.8% 11.1% 15.1%
Note: the sectors are grouped according to the classification adopted in table 2 (see “Literature). Labour productivity is presented in levels.
31 of all three sectors for both periods 1995-2001 and 2002-2008. A reason for this could be the replacement of workers by machines (Maddision, 2007). In addition, industry and services show a higher labour productivity than the agricultural sector. Therefore, the shift from worker out of agriculture into industry and services leads to an increase in overall labour productivity.
In table 7 the (partial) factor contribution of the total economy and the three aggregated sectors are presented for all three periods. It shows the difference (1) value added coefficients , (2) labour inputs , (3) shares of sectoral demands that are fulfilled domestically , (4) input mix , (5) the intra-sectoral shares of final demand and, (6) the inter-sectoral mix of final demand . By taking the natural log of factor values we captured factors’ importance for effecting change in labour productivity. A value larger than 0 implies that this factor makes a positive contribution to the growth rate of labour productivity. When the value is below 0, the factor makes a negative contribution to labour productivity. The decomposition results reveal some general insights on the study periods. It shows that labour savings (i.e. less labour inputs per unit of output) have consistently dominated across all periods, both for the total economy as well as the three aggregate sectors. Partial factor has had the largest positive effect on labour productivity throughout the years.
32 D) Services Factor V E P A B Y 1995-2001 6.15% 120.24% 0.86% -9.97% -14.60% -2.68% 2002-2007 1.45% 98.34% -0.69% -11.95% 14.74% -1.90% 2002-2008 1.06% 100.56% -0.49% -10.40% 10.26% -0.99%
Note: the abbreviations represent the factors contributing to annual labour productivity growth: V, value added per unit of output; E, labour input per unit of output; P, domestic supply ratio; A, technological change; B, intrasectoral shares; intersectoral mix of final demand y.
Looking at the total economy, the contribution of labour savings E have slightly declined comparing 1995-2001 with the period from 2002-2007, from 97.16 percent to 95.68 percent (table 7). However, when considering period 2002-2008, the contribution of labour savings has actually increased slightly. Considering the period 1995-2007, the results seem to be in accordance with the findings of Yang and Lahr (2010), who also found a decrease in the contribution of labour savings.
Considering the three aggregate industries, we find that industry experienced the most benefits from labour savings, with 106.84 percent, 133.66 percent and 124.72 percent throughout the three periods (table 7). Comparing period 2002-2007 and 2002-2008, we see the contribution of labour savings have slightly decreased.
Table 8: Factor contribution per subsector of the industry sector, textiles and food (1I), wood and paper (2I), petroleum, chemicals and metals (I3), machinery and electronics (I4) and manufacturing and construction (I5) (%)
33 C) 2002-2008 Index V E P A B Y 1I -36.21% 132.90% -1.91% 3.87% 8.39% -7.05% 2I -106.51% 209.12% 0.30% 0.82% -1.84% -1.88% 3I -32.62% 127.50% -1.22% 4.34% 1.38% 0.62% 4I -79.18% 178.41% 0.35% -0.05% -0.15% 0.60% 5I -24.36% 112.90% -2.22% 11.92% 4.07% -2.30%
Note: the abbreviations represent the factors contributing to annual labour productivity growth: V, value added per unit of output; E, labour input per unit of output; P, domestic supply ratio; A, technological change; B, intrasectoral shares; intersectoral mix of final demand y.
34 measures that liberalized the labour market – such as the adoption of the liberalizing labour law in 1986 (see “Literature”).
In table 9, the results of the decomposition analysis for the service sectors are presented for each period. In the first period we see that 1S, 3S and 4S experience a positive influence of V (i.e. value added per unit of output), which indicates that China has been able to obtain more value added per unit of output during this period in these industries. However, in the subsequent periods the contribution of V turns negative, for sectors 1S: trade, and 4S: government services. Sector 3S experienced a slight decrease however remains positive. In addition, 2S becomes positive in the second and third period.
Table 9: Factor contribution per subsector of the services sector, trade (1S), transport (2S), communication, finance and real estate (3S), government services (4S), other services (5S) (%) A) 1995-2001 Index V E P A B Y 1S 7.61% 79.68% 0.68% 13.41% -8.64% 7.27% 2S -11.24% 128.69% 0.04% -7.20% -9.91% -0.39% 3S 7.76% 114.03% -2.19% -14.68% -2.12% -2.80% 4S 7.78% 92.38% 0.06% 0.10% -0.39% 0.07% 5S -6.57% 160.20% 1.56% -11.19% -33.92% -10.09% B) 2002-2007 Index V E P A B Y 1S -1.11% 94.19% -0.29% 16.49% -1.74% -7.53% 2S 6.40% 130.37% 0.02% -38.86% -0.34% 2.41% 3S 5.58% 95.85% 0.52% -2.08% 3.02% -2.90% 4S -12.73% 112.74% 0.32% 0.54% -0.94% 0.07% 5S -8.65% 120.62% -1.41% -23.16% 9.92% 2.69% C) 2002-2008 Index V E P A B Y 1S -0.66% 97.59% -0.32% 10.68% -1.72% -5.57% 2S 4.81% 122.63% 0.22% -29.61% 0.39% 1.56% 3S 4.33% 99.18% 0.45% -3.28% 1.09% -1.76% 4S -9.63% 110.88% 0.25% -0.97% -0.57% 0.04% 5S -6.70% 119.52% -1.08% -20.97% 6.48% 2.75%
35
technological change; B, intrasectoral shares; intersectoral mix of final demand y.
For partial factor E, we again see major positive contributions. 1S, 2S and 4S show an increase in the contribution of E when comparing the first and second period. 3S and 5S show the opposite trend. The highest contribution is found in 2S with the top percentage of 130.37 in the second period.
Concerning partial factor P, we see that the contribution is only minor and both positive as well as negative. It seems that the domestic supply ratio is of low influence on the services sectors. This makes sense as there will not be major shifts in the domestic supply ratios for services, due to the fact that services generally don’t have much foreign intermediate inputs.
Lastly, one thing that stands out in the table is that partial factor B shows only negative contribution values for the first period, which turns around to some positive values for the second and third period. The influence of B has thus become more positive in general.
Table 10: Factor contribution per subsector of the industry sector, second division: market services (MS) and government services (GS) (%)
A) 1995-2001 Index V E P A B Y MS 0.63% 99.81% 0.57% -5.62% 0.17% 4.43% GS 6.53% 99.49% 0.70% -6.35% -8.75% 8.39% B) 2002-2007 Index V E P A B Y MS 2.70% 87.12% -0.18% 8.77% 6.61% -5.02% GS -8.37% 107.89% -0.63% -13.73% 26.79% -11.95% C) 2002-2008 Index V E P A B Y MS 2.04% 93.94% -0.07% 4.98% 2.69% -3.58% GS -6.64% 106.25% -0.48% -11.95% 21.58% -8.77%
Note: the abbreviations represent the factors contributing to annual labour productivity growth: V, value added per unit of output; E, labour input per unit of output; P, domestic supply ratio; A, technological change; B, intrasectoral shares; intersectoral mix of final demand y.
36 been able to obtain more value added per unit of output during this period in these industries. We observe a significant increase from 1995-2001 to 2002-2007. However, when comparing the second to the third period, we observe a slight decrease in V, which indicates that China has lost some value added per unit of output. When looking at partial factor E for MS, we see a slight decrease in contribution for the second period. However, comparing the results to the third period we see a higher contribution. In addition, partial factor P turned negative after the first period. Considering partial factor A for sector MS, we see that the contribution turned positive after the first period. For sector GS, factor A shows a negative contribution after the first period, which becomes a little less negative in 2002-2008. Subsequently, we look at partial factor B. For GS it shows a negative contribution in the first period, where it has a rather large positive contribution in the later two periods. Lastly, partial factor Y turns negative after the first period.
Table 11: Annual labour productivity growth rates and annual export growth rates for all 33 industries, per period (%)
A) Period 1995-2001
1995-2001
Sector Industry Labour Productivity Exports
37 I 19 -4% 3% I 17t18 5% 3% I E 9% 2% S L 15% 2% I 15t16 5% 2% S N 15% 1% I F 2% 1% S M 13% 1% S H 7% 0% S 70 4% 0% I 20 3% -1% A AtB 2% -2% S J 2% -4% S 63 -2% -7%
Note: the abbreviations represent the sector classification based on an aggregation level of three industries: A agriculture, I industry and S services.
B) Period 2002-2007
2002-2007
Sector Industry Labour Productivity Exports
38 S M 11% 7% S 60 6% 7% S 52 15% 6% S 51 15% 6% I E 9% 6% A AtB 12% 6% I 23 0% 6% S L 11% 5% I 21t22 0% 5% A C 12% 2% S 70 18% 0% S O 16% -2%
Note: the abbreviations represent the sector classification based on an aggregation level of three industries: A agriculture, I industry and S services.
C) Period 2002-2008
2002-2008
Sector Industry Labour Productivity Exports
39 S 51 17% 8% I E 11% 8% S L 13% 7% I 21t22 3% 6% A AtB 14% 6% A C 13% 5% S 70 17% 0% S O 17% 0%
Note: the abbreviations represent the sector classification based on an aggregation level of three industries: A agriculture, I industry and S services.
Table 11 shows the labour productivity growth rates and export growth rates for all sectors per period. As we can see, the first period shows somewhat higher labour productivity growth rates on top – with the higher export growth rates. However, when looking at the second and third period, there appears to be no relation whatsoever between the labour productivity growth rates and the export growth rates.
6. Conclusion
Based on the results we can answer the research question which sectors are driving China’s aggregate economic growth during 1995-2008 and which factors are driving the growth of these sectors?” and see whether our expectations match the outcomes.
40 which implies less labour is involved in the agricultural sector in the periods 2002-2007/8 than in 1995-2002. As expected, these labour inputs move to industry, the sector that shows the largest increase in labour inputs. Lastly, services shows a decline in labour inputs, however the total labour productivity increases (table 7) which could for instance indicate an increase in the adoption of technology in this sector. Based on these findings, we can adopt hypothesis I and conclude that structural change significantly contributes to the increase in labour productivity growth and hence to China’s aggregate growth pattern.
Largely due to the fact that China possesses such an enormous labour force, it has been able to take the advantages of structural change (McMillan, Rodrik, 2011). In 1979, just a few years after Mao Zedong’s death in 1976, the country entered a period of extensive economic reforms. The main aim of these reforms was to achieve an annual GDP growth of at least 8 percent, partly in order to create enough jobs to keep the Chinese people satisfied and prevent possible uprisings (Yang, Lahr, 2010). One of these reforms was the liberalization of the labour market – mainly through the implementation of a more liberal labour law (see “Literature”) (Yang and Lahr, 2010). This new law has led to a better allocation of labour across sectors. This is because workers can switch easier, which leads to the workers to be in a more effective and labour productive spot. Because of this, the workers become more productive and thus partial factor (i.e. labour inputs per unit of output) decreased in absolute terms. This causes the labour productivity to grow. As has been pointed out in the literature, Yang and Lahr (2010) argue that the Chinese labour market has experienced an increased flexibility. As the more flexible labour market should lead to a better allocation of labour and therefore a more efficient labour use, we expected that the labour use per unit of output will decrease. A decrease in labour use per unit of output will be shown in the form of a positive contribution of E. Focusing firstly on the aggregated sector (table 8), we see that for all sectors E indeed has had a major positive contribution. This is in great support of the hypothesis. When looking at the disaggregated industries, the same picture can be seen. All industries show a major positive contribution in E, therefore hypothesis 2 is considered supported.
41 China even managed to achieve an annual growth rate of 10.4 percent (Worldbank, 2014). Between 1979 and 2009 China’s exports have grown by an impressive 16 percent annually (Lin, 2010). Currently, China is considered world’s largest exporter of goods, with 10.2 percent of the global share (World Bank, 2014). As has been explained in the literature, we hypothesized that the export intensive industries in Chinese economy, which are highly exposed to international competition, have contributed significantly to China’s aggregate productivity growth. In order to test whether this has been the case for China, we have composed table 11. In this table, we show the labour productivity growth levels as well as the export growth levels for all industries. The levels have been subsequently ranked from highest to lowest, based on the export growth levels. Following the line of reasoning from Yang and Lahr (2010), we expect the industries on top to have the highest labour productivity growth. As can be seen for the first period, the industries with the biggest growth in exports show a somewhat higher labour productivity growth as well. When looking at the second and third period however, we see that the labour productivity growth is not at all bigger for the industries with the highest growth in exports. Based on these results, we reject hypothesis III.
Over the years, China’s income per capita has risen to a level of a low-income country to a well-established middle-income country. Its per capita income as a percentage of the United States has increased from 1.6% in 1980 to 10.9% in 2012 (World Bank, 2014). The great development that China has experienced provides many interesting aspects to investigate. This thesis analysed the structural change that China has experienced. It has posed three hypotheses. Hypotheses I and II have been supported by the results found in this research. For hypothesis III, this thesis did not find any support, coming back to the research question of this thesis:
Which sectors are driving China’s aggregate economic growth during 1995- 2008 and which factors are driving the growth of these sectors?
42 second and third period. Concerning the factors, there is no doubt about which one is driving the growth for the labour productivity in China as well as for all sectors individually. Partial factor E has contributed by far the most to the labour productivity growth in all sectors. The decrease in the amount of labour inputs per total outputs has been the major cause of the labour productivity growth. In addition, input mix and the intra-sectoral shares of final demand have had significant influence.
7. Discussion
In this thesis we have investigated what sectors drive the Chinese economic growth, and what partial factors foster this growth. As has been explained, we have used a structural decomposition analysis based on the article of Yang and Lahr (2010), who have based their method on Jacob’ (2003). The reason for using a decomposition analysis has been to capture the inter-industry linkages and to include both the demand and the supply side of the story; something that cannot be achieved by using growth accounting. During this research we came across a mistake in the notation of one of the formulas in Yang and Lahr (2010), which is represented by formula 1.2 in our research (see “Methodology”), which represents the changes in labour inputs.
(1.2)
changes in labour inputs
Here the formula starts with component , which represents the labour inputs per unit of output. In the article of Yang and Lahr (2010), the terms and were swapped. For our research the terms are exchanged, represented by equation 1.2 presented above.
Besides, we will highlight some additional findings of this research. First of all, when considering the total economy in table 7, the contribution of labour savings
has slightly declined comparing 1995-2001 with the period from 2002-2007, from
43 period 1995-2007, the results seem to be in accordance with the findings of Yang and Lahr (2010), who also found a decrease in the contribution of labour savings. However, when including the year 2008 into the formula we see that the contribution of labour savings is higher in the period 2002-2007 than 2002-2008. The increase due to the addition of the year 2008 is therefore interesting. What triggers this increase, will this trend last and what does this imply for the future? Further research should be done to be able to answer these questions. For now we will pose some suggestions. For instance, perhaps this indicates that growth of the economic growth rates is decreasing. As China industrialized rapidly, making the total industry much more efficient, there will be a point at which the growth rate will decrease. Possibly, this this downwards turn has been set in in the second half of the 2000s, which is represented by the results in the year 2008 of our research. Another cause could be the global financial crisis of 2008. A crisis could for instance lead to a decrease in contribution of other factors such as the inter-sectoral mix of final demand (y), which in turn would lead to a relative increase of contribution of labour savings. The same trend can be seen in multiple disaggregated sectors as well. The differences in results for 2007 and 2008 support our decision for including both years into our analysis. Secondly, the negative effect of value added per unit of output V is interesting (table 7). The negative V could be an indication that the opening up to trade that has occurred in China, has led to a much more competitive environment for the Chinese companies. In this competitive environment, prices have to be lowered in order to stay competitive, which leads to lower profit margins and perhaps lower salaries. As these two aspects decline, the value added for China decreases, which would explain the negative contribution of V. Nevertheless, these suggestions should be examined before we can make solid assumptions.
