The Scale of Government Procurement in China Negatively Affects. the Purchasing Performance: Empirical Analysis Based on Tobit

Transcription

1 The Scale of Government Procurement in China Negatively Affects the Purchasing Performance: Empirical Analysis Based on Tobit Model of Provincial Panel Data Bai Zhiyuan (School of Public Finance and Taxation, Zhongnan University of Economics and Law) [Keywords]Government Procurement Performance, Government Procurement Effective Scale, Public Finance, Fiscal Expenditure I. Introduction Government procurement performance is the quantitative evaluation, which is based on centralized procurement rules, of procurement functions and objectives. From the aspect of saving funds, the performance of government procurement in China is outstanding. Over the past decade, since the implementation of the Government Procurement Law, a total of over 660 billion yuan of financial funds have been saved. The current government procurement scale is still huge. In 2014, the government procurement scale has reached 1.73 trillion yuan, accounting for 11.4% of the total fiscal expenditure, accounting for 2.7% of GDP [1]. It is noteworthy that whether such a large-scale government procurement has produced a positive impact on purchasing performance. At present, the academic circles mainly study government procurement performance from two aspects. First, from the perspective of research, the two most basic metrics of purchasing performance are efficiency and effectiveness in order to meet the procurement goals and increase the procurement value. Efficiency measures how the input is transformed into output successfully, and effectiveness measures how it succeeds in achieving its desired output. Based on the contrast relationship between output and input, scholars have used comprehensive assessment [2] and specific quantitative assessment to determine the achievement of established goals and responsibilities [3], the government procurement capacity, the performance and effectiveness of the procurement participants, [4] to evaluate the functional goals of government procurement policies and the realization of economic effectiveness goals [5], to reflect and improve the level of procurement. [6] Secondly, based on the research methods and perspectives, scholars have studied

2 the positive correlation between the purchasing maturity (strategy, electronics, process, information, supervision and organization) and purchasing performance; [7] have studied the importance ordering of procurement performance factors, which include employee competitive variables, organizational resources, the impact of stakeholders, and government policies. [8] Some scholars also come to different conclusions through the empirical research of the Agency for Retirement Benefits, and the order is procurement policy, information technology, staff training and top-level support. [9] It can be seen that at present the academic community has a profound understanding of government procurement performance and is fully aware of the importance of analyzing the impact of government procurement performance and strengthening the performance management of government procurement. For example, the measurement of performance includes the measurement of output and result, so that the output performance index is the most suitable in government procurement. [10] The assessment system includes the assessment of the target duties and responsibilities of regulators, purchasers and agencies and the assessment of information disclosure. [11] The analytic hierarchy process (AHP) is used to determine the weight of government procurement performance evaluation indicators based on public value. [12] However, there is little research literature on the correlation between government procurement scale and procurement performance, which is also the original intention of this paper. In theory, as the largest purchaser of goods, construction and services on the market, the government procurement has a significant impact on the market prices. With the formation of a government procurement buyer's market, government departments can always buy "good quality" goods, construction and services at lower prices. The scale effect formed by centralized purchasing greatly reduces the purchasing cost and transaction cost, and helps to open up the use of the project budget and reduce the unnecessary duplicated procurement, thus saving the budgetary fund. This is an important manifestation of the government procurement function of saving funds. If we see the government procurement as a whole mechanism, its resource allocation can take the common goal and the overall planning as the core, connecting different buyers through a certain procurement organization, thus forming huge scale effect that is required by policy function of government procurement. Therefore, the appropriate scale of procurement is the key factors of

3 government procurement performance to achieve economies of scale, fully exert the government procurement function, and improve government procurement performance. This paper tries to test this theoretical presumption. We use government procurement scale as the main explanatory variable to study government procurement performance, based on the data of 31 provinces and autonomous regions in China from 2007 to 2014, we construct the panel Tobit model, and make an empirical analysis of the relationship between government procurement scale and government procurement performance, and put forward the corresponding policy recommendations. II.The Model and Measurement of Government Procurement Scale Affecting Procurement Performance As the government procurement system in China only lasts for more than ten years, the relevant performance index data are not comprehensive, which is also an important reason that makes the evaluation of government procurement performance difficult to carry out in depth. The evaluation of government procurement performance is determined by the evaluation index and the weight of each index. No matter which index system is adopted, it must include indicators such as the government procurement scale, the savings rate, etc. Therefore, although it is impossible to conduct a comprehensive assessment of government procurement performance, government procurement scale can be used as a measure of procurement performance to analyze the factors influencing government procurement performance. (a) Model Construction and Index Measurement As mentioned above, in addition to government procurement scale, government functions have an important influence on the performance of government procurement. Government functions are closely related to economic development. Therefore, the level of economic development is taken as an alternative variable for government functions. Not only that, the level of economic development also directly affects the level of government financial resources. At the same time, given the profound influence of fiscal decentralization on the scale of central and local procurement, it is necessary to study the impact of fiscal decentralization on the performance of government procurement. Therefore, this paper establishes the

4 following panel model: jzlit=β1cggmit+β2lgdpit+ β3openit+ β4fisit+ αi+ ηt+ μit Based on the availability of data, the sample period for panel data analysis in this paper is from 2010 to Since the implementation of the "Government Procurement Law," China has saved a total of more than 660 billion yuan in financial resources. In 2012, it saved billion yuan in financial capital and the saving rate is 11.7%. [13] The saving rate of government procurement is an important factor in evaluating government procurement performance. Therefore, we use the saving rate (jzlit) as a measure of government procurement performance and use it as an explained variable. The scale of government procurement (cggmit), measured by the proportion of government procurement budget to fiscal expenditure, is the key controlled variable in this paper. αi ηt are the fixed effect and year fixed effect respectively, μit is the error item. According to the existing literature and the above theoretical analysis, this article also controls the following economic factors. (1) the level of economic development (lpgdpit), with the description of natural logarithm of real GDP per capita; (2) the degree of opening up (openit), which is measured by the ratio of import and export to GDP;(3) The degree of fiscal decentralization(fisit), measured by the per capita fiscal expenditure / (per capita fiscal expenditure + central fiscal expenditure). Saving rate (jzlit) Variable Government procurement scale (cggmit) Per capita GDP (lpgdpit) The degree of opening up (openit) Table 1 Descriptive Statistics of Major Variables Mean Standard deviatio n Minimu m Maximum Sample size overall N =131 Between groups n =31 In groups T =5 overall N =155 Between groups n = 31 In groups T =5 overall N =155 Between groups n = 31 In groups T =5 overall N = Between n =31 groups

5 The degree of fiscal decentralizat -ion (fisit) Based on the analysis of government procurement performance in various provinces, we find that during the sample period, the performance of government procurement in the eastern region is significantly higher than that in the central and western regions, but there are significant differences among the provinces. The performance of some provinces are relatively stable, such as Shaanxi and Hunan; while some provinces show a downward trend, such as Guangxi, Jiangxi; some provinces are on the rise, such as Hainan, Yunnan; some provinces have large fluctuations, such as Tianjin, Fujian. To some extent, the inter-provincial differences in procurement performance can help us estimate the impact factor of government procurement performance. (b) Regression Analysis of Initial Model As the explained variable, the government procurement performance (jzlit) is a relative number between 0 and 1, which belongs to the "restricted explained variable", the direct OLS estimation is both biased and inconsistent. So that we need to use the panel Tobit model. This section first uses the cross-section regression and mixed regression (POLS) to do Tobit regression analysis, the results in Table 2. Table 2 Empirical Results of the Influencing Factors of Government Procurement Estimated parameter In groups T = 5 overall N =155 Between groups model 1 (Cross Section Regression) *** (-3.13) *** (3.54) * (1.87) *** (6.49) *** (7.55) Performance: Initial Mode Model 2 (POLS) n =31 In groups T =5 model 3 (POLS-robust) model 4 (Random Effect) model 5 (Time dummy variable) β * ** (-1.94) (-2.09) (-1.17) (-1.120) β *** *** ** * (3.25) (4.09) (2.18) (1.740) β ** *** * (2.27) (3.07) (1.23) (1.885) β *** *** ** *** (-2.45) (-2.92) ( ) (-2.771) Constant *** *** *** *** term (5.12) (6.09) (8.46) (2.994) Year *

6 (-1.676) Year *** ( ) Year ** (1.956) Year (-1.149) Adj R Log likelihood σ u σ e ρ Note: *** ** * respectively means significant at 1%, 5% and 10% significance level; The data in parentheses in the table is the z statistic of each parameter. The model 1 is based on the cross-sectional Tobit regression analysis of crosssectional data of 31 provinces in China in The results show that the impact of government procurement scale and fiscal decentralization on government procurement performance is significantly negative. The government procurement scale increases by 1%, and government procurement performance decreases 8.83%; fiscal decentralization increases by 1%, the procurement performance decreases 8.61%. However, the level of economic development and the degree of opening up have a significant positive impact on the performance of government procurement. For every 1% increase in per capita income, the performance of government procurement increases by 1.32%. The degree of opening increases by 1%, the performance of government procurement improves by 4.08%. However, crosssectional data only contains information for the year of 2014, so the Model 2 does a Tobit regression analysis of mixed data that included data from 2007 to We finds that the regression results show a significant change compared to Model 1. The coefficient of economic development is only half of the original, and the impact of fiscal decentralization on the performance of government procurement has significantly diminished. On the level of significance, the coefficient of opening up is even more significant, but the coefficient of the procurement scale is less significant. Different standard deviations assume that perturbation terms of different periods are irrelevant, but in fact, the disturbances in different periods of the same individual are often auto correlated. Thus, we should use the clustering robust standard deviation to estimate standard deviation. Therefore, Model 3 uses the estimate standard deviation and finds that the coefficient of each parameter changes little, and the effect of government procurement on procurement performance is significantly negative, but

7 β1 β2 β3 β4 the clustering robust standard deviation is greatly reduced compared with ordinary standard deviation, which makes t-statistics larger and increases the significance of model coefficients. Due to ignoring the influence of "limited explanatory variables" on the unbiasedness and consistency of the model and treating it as normal data, the regression results of Model 4 and Model 5 show that compared with the first three models, there is a big change in the coefficient, and the significance is not well enough. (c) Regression Analysis of Panel Tobit Model In order to use the cross-section and time series information included by panel data and avoid the bias caused by the constrained OLS estimation, the following is a regression analysis of the Tobit random effects panel model with constrained variables. Table 3 Empirical Results of the Influencing Factors of Government Procurement Performance: Panel Tobit model Estimated parameter Model 6 (nation) *** (-3.26) Constant term *** (11.21) Standard deviation of individual effect standard deviation of interference term Model 7 (nation) ** (-2.22) *** (2.27) *** (10.48) Model 8 (nation) * (-1.77) *** (3.26) ** (2.34) *** (9.70) Model 9 (nation) * (-1.89) *** (3.32) ** (2.22) *** (-3.72) *** (9.88) Model1 0 (east) (1.32) * * (2.19) * * (2.30) (1.23) * * (2.27) Model11 (Central) * (-1.69) (1.04) * (1.76) ** (-2.01) *** (7.51) Model12 (west) *** (-3.08) *** (2.91) ** (-2.15) *** (-2.79) *** (4.99) Likelihood ratio testχ ρ Log likelihood Note: *** ** * respectively means significant at 1%, 5% and 10% significance level; The data in parentheses in the table is the z statistic of each parameter. Model 6 to Model 9 are based on the national data, and Model 6 only includes one explanatory variable, that is government procurement scale. What s more, estimated results show that the coefficients are significantly negative, so if the amount

8 of government procurement budget accounting for the proportion of financial expenditure increases by 1%, government procurement performance will drop by 5.28%. However, this estimate has missed the key explanatory variables and led to skewed estimates. So, Model successively add new control variables for robustness testing. The regression results show that, firstly, although the number of control variables have been increasing, the impact of government procurement scale on government procurement performance has been significantly negative. Secondly, although the regression conclusions from Model 1 to Model 5 all indicate that the scale of government procurement has a negative effect on improving government procurement performance, the coefficients between Table 2 and Table 3 reveal that due to limitations of data information and the impact of restricted variables, Model 1 to Model 5 overestimated this negative effect. Thirdly, the level of economic development and the development of opening up have a positive effect on promoting government procurement performance, while the increase of fiscal decentralization has a negative impact on the performance of government procurement. China has a vast territory and a wide range of regional development differences, so if we do the regression with data at the national level, we will ignore the differences between regions. Therefore, Model 10 to Model 12 are regressed based on sub-region data, and we find that, firstly, the impact of government procurement on government procurement performance shows obvious regional differences. The central and western regions are significantly negative while the eastern regions, though negative, are not significant. Secondly, the impact of economic openness on the performance of government procurement also has geographical differences. It is significantly positive in the eastern and central regions but significantly negative in the western regions. Thirdly, the impact of fiscal decentralization on government procurement performance is significantly negative in the central and western regions, which is consistent with the regression results based on national data, but it is not significant in the east. III. A General Explanation of the Government Procurement Scale Negatively Influence the Performance of the Procurement In this paper, the saving rate is the main measure of government procurement performance, and we build a panel Tobit model based on data from 31 provinces and autonomous regions in China from 2007 to 2014 to make an empirical analysis of the relationship between government procurement scale and government procurement

9 performance. The results indicate that with the continuous expansion of the scale of government procurement, the performance does not increase; what s more, procurement scale has a significantly negative influence on procurement performance. After joining the control variables such as the level of economic development, the degree of opening up and the fiscal decentralization, the conclusion remains stable. After analyzing reasons behind, we find that: Firstly, the concept of government procurement system lags behind. The absolute size of China's government procurement increases from billion yuan in 2002 to 1.73 trillion yuan in The proportion of absolute size of government procurement to fiscal expenditure increases from less than 5% in 2002 to 11.4% in 2014; and the proportion of absolute size of government procurement to GDP increases from less than 1% in 2002 to 2.7% in [14] Compared to the participants of Government Procurement Agreement (GPA) and international standard that government procurement accounts for 10% of GDP and 30%-50% of government fiscal expenditure, China's government procurement is characterized by its small size and narrow implementation scope and so on. At present, government procurement in China is mainly limited to goods and a small amount of service, and many procurement funds are still free from government procurement regulation. The proportion of procurement to GDP is still very low, and its policy performance is not very satisfactory. By 2012, only 8 provinces had statistics on the policy function of government procurement, which was the same as [15] From the existing data analysis, under the condition of the small scale of government procurement in China, the scale of procurement, which plays the role of policy function is less. Therefore, the focus of China's existing government procurement system is saving and preserving fiscal funds. The scope of regulation of government procurement system is limited to the procurement activities using fiscal funds. The position of procurement policy from the design of government procurement system is not appropriate, resulting in a narrow range of government procurement and procurement regulations, which cannot meet the functional requirements of the government procurement performance. Secondly, the coverage of government procurement law is not coordinated. The procurement of goods had the absolute advantage in 2003, while engineering procurement did in In 2014, the purchase amount of engineering projects was 1.01trillion yuan, accounting for 58.6% of the total procurement. [16] On the other hand,

10 because of small-caliber on the scale of engineering procurement in China, which means the bidding is suitable for the "Bidding Law", the engineering procurement is not within the scope of the government procurement statistics. Even if the largecaliber statistics shows that the proportion of engineering procurement is dominant, engineering procurement is not within the scope of the government procurement law. In fact, the scale of government procurement does not include the scale of engineering procurement. When the government procurement of goods, engineering and services apply the "Bidding Law", not only does the "Bidding Law" that covers all the bidding procedures overlap the "Government Procurement Law" that covers government procurement, but also some ambiguous boundaries exist. This provides the legal basis for engineering procurement to be no longer subject to the government procurement system. In addition, there exists the probability of using inconsistencies on the coverage of "Government Procurement Law" and "Bidding Law" to circumvent government procurement, thus further reducing the scale of government procurement. Thirdly, the technical procedures of government procurement is unscientific. As for procurement method, in 2013, the scale of open bidding was 1.36 trillion yuan, accounting for 83.3% of the total procurement volume. The method of open bidding still holds an absolute advantage. [17] Over the years, open bidding has been dominant in government procurement, which indeed plays an important role in regulating the government procurement behavior. However, there are also shortcomings of too much rigidity and lack of flexibility. Open bidding places too much emphasis on program management, which increases procurement management costs. Therefore, the existing government procurement system emphasizes open bidding. At the same time, other methods of procurement have strict implementation and evaluation criteria in the technology and procedures. It lacks flexibility to meet policy goals. The methods and technical procedures of government procurement do not support the implementation of procurement policies, thus reducing the scale that government procurement should own. In addition, since the 18th CPC National Congress, the government has greatly increased the amount of services purchased from social forces, and the service procurement has been growing rapidly. However, the purchase of government services belongs to the management of the General Financial Sector, and does not belong to the supervision of the Government Procurement Supervision Department, which is not conducive to coordination and cooperation in policy and

11 methods. Therefore, in practice, government procurement methods and technologies are not be used effectively, and most of them are out of government procurement regulations. The inappropriate methods and procedures also weakened the scale of government procurement. Fourthly, the professional purchasing ability is not strong. As the employees of centralized procurement agencies in China are mainly from administrative units, the expertise and technology are different; the lack of professional ability makes it very difficult for government procurement to play a role in achieving policy goals. Social agencies do not have a sense of responsibility of government procurement policies, while the purchasing ability is less capable under the implementation of decentralized procurement policy. Therefore, the lack of professional ability further reduces the scale of government procurement. VI. Policy Suggestions on Realizing the Positive Scale Effect of Government Procurement At present, China is in a critical period of deepening the reform in an all-round way and accelerating the transformation of the mode of economic development. Therefore, giving full play to the policy functions of government procurement can contribute to the realization of these strategic objectives. From the central government to local governments, all are actively promoting the reform and innovation of the government procurement system. However, the performance of government procurement in China is still not very satisfactory. In order to improve the performance of government procurement furtherly, in addition to promoting economic development and rational fiscal decentralization, it is necessary to further expand the absolute and relative size of government procurement scale and promote the realization of the positive scale effect of government procurement. Specifically: Firstly, the government should reasonably locate the concept of government procurement system, redefine the connotation and extension of government procurement and expand the scope of government procurement regulation. Government procurement is one of the powerful tools for achieving government goals. Government procurement strategy is an integral part of government strategy. The system concept of government procurement to save funds and prevent corruption is outdated. Emphasis on "value for money" is the maximum benefit that the government can make when it considers all the costs. The government procurement system should be reformed to emphasize the realization of policy functions. According

12 to the legal definition of government procurement in China, the scope of government procurement management has great limitations in terms of the nature of funds, the content of procurement or the amount of procurement. We need to rethink the meaning of government procurement and expand the scope of government procurement management, which also lead to the expansion of government procurement scale. Depending on the nature of the funds, the scope of government procurement market will be expanded from the current use of financial funds to public procurement, including the procurement follow-up of the use of international loan funds, donations and financial aid These funds are ultimately financed by financial funds or its user is the entity that implement the will of government. According to the procurement content, the content should be expanded from the scope of procurement included in the centralized procurement catalog to the government procurement outside the catalog. According to the standard of the amount of purchasing money, the procurement should change from a certain quota standard to a penny of purchase; that is, the scale of government procurement is not only the scale of centralized procurement, but also include all the scale of decentralized procurement. Though the formation of a unified centralized management, although the decentralized procurement appears to be fragmented, the actual is not dispersed, and with the centralized procurement, it results in "synergy" scale effect. Depending on the purpose of the procurement, the scope of the government procurement regulation will be expanded to the area of procurement for the realization of the will of the government, which will help fulfil its obligations after joining the Agreement on Government Procurement (GPA) and facilitate the protection of domestic industry and policy goals. We look forward to establish a unified large market for public procurement, which help to coordinate the government procurement policies and systems covering government procurement, engineering procurement, state-owned enterprises procurement and military procurement. Secondly, the government should coordinate the conflicts of the two laws covering government procurement. Although the proportion of public investment expenditures (including infrastructure expenditures and non-productive infrastructure expenditures) in the composition of fiscal expenditure fluctuates periodically, the proportion is relatively stable. The scale and proportion of engineering procurement also increase and are relatively stable. At the same time, the proportion of China's public economic sector is relatively large. Some of the state-owned enterprises still

13 implement the government's will. What s more, the cost of PPP projects vigorously implemented in various places is ultimately borne by taxpayers' funds. After joining GPA, PPP projects will be subject to GPA regulation. These entities and projects should be included in the regulation of government procurement, or else we will lose some "market-changing market" opportunities and exchange of bargaining chips. The conflict between the coverage of the two laws should be coordinated, and government procurement projects, some state-owned enterprises and PPP projects should be included in government procurement regulations to achieve the positive scale effect of the existing government procurement. Thirdly, the government should increase the flexibility of technology and procedures. Open tendering is not omnipotent; it is just a "tool" in purchasing "toolboxes", though it may be the more important one. The use of open bidding or any kind of procurement tool does not necessarily lead to the automatic realization of the goals of saving and increasing efficiency, fighting corruption and promoting government policies and functions. With the change of procurement objects (from engineering, goods to service-oriented), the change of procurement value concept (from the pursuit of a single cost-effective ratio to the pursuit of the unification of economic benefits and social benefits), the enhancement of purchasing staff's ability (from a compliance procurement to a reasonable and efficient procurement), and the improvement of procurement environment, (which is composed of the purchasers, suppliers, reviewers, procurement agencies, regulatory agencies, the public and society), a variety of non-tender procurement will have more room for development. The government should further strengthen the application of competitive negotiation methods in actual procurement, weaken the leading position of open tendering and enhance their adaptability to the policy functions of government procurement in the existing procurement methods and procedures. At the same time, the general financial department and the government procurement supervision department need to make overall plans, clarify the power and responsibilities, focus on the business, adopt more appropriate procurement methods and procedures in purchasing services, especially realize the policy functions of government procurement in purchasing service and improve government procurement performance. Fourthly, the government should improve the ability of professionals to achieve government procurement goals and further improve the efficiency of government procurement. To achieve the government's wishes and goals through government

14 procurement requires the necessary ability of professionals. These people need outstanding communication skills because they have to be good writers, negotiators, speakers, market developers and consumer service agents; they need strong adaptability, managerial skills and teamwork; they also need professional expertise of government procurement, including competitive negotiation techniques and capabilities. Through the qualification examination, regular training mechanism and promotion mechanism to promote the improvement of government procurement professionals, and promote the scale effect of government procurement policies. It is also the primary condition to improve the performance of government procurement. In short, based on the principle that the government leads the market and maximizes the efficiency of government resources allocation, and based on the effective policy function of government procurement, we should update the concept of government procurement system. Within the scope of broad government procurement defined by the public interest, we should emphasize the input and output of government procurement policies, adopt flexible and diversified procurement methods based on competitive negotiation, improve the ability of government procurement professionals to implement government procurement policy comprehensively), and make use of the Government Procurement Agreement to extend the international application scope of the domestic procurement policy by exchanging markets. Over all, it is necessary to ensure that the scale of government procurement has a real positive impact on the performance of government procurement, and government procurement can give full play to its policy functions under the new normal. References: [1] [2] Zhou Meng. The establishment and improvement of government procurement performance evaluation system [J]. Fiscal Supervision, 2012 (30). [3] PPDA. Report on Compliance Checks undertaken on 120 Procuring and Disposing Entities[EB/OL]. rtaken.pdf/, [4] Gao Zhiyuan. International Procurement Performance Evaluation of Government Procurement [J]. China Procurement, 2013 (9). [5] Xu Bei. Evaluation Model of Sunshine Index of Government Procurement Based on Fuzzy

15 Analytic Hierarchy Process [J]. China Government Procurement, 2013 (7). [6] Martinez, J. G. Effective ERP and Supplier Coordination for Procurement Performance: A cross national study. Inter Metro Business Journal, 2008(4). [7] Batenburg. Procurement Maturity, Alignment and Performance: a Dutch Hospital Case Comparison. 22nd Bled Conference Enablement: Facilitating an Open, Effective and Representative society, June 14-17, [8] Kavua, K. B.,and Ngugi, K. Determinants of Procurement Performance of Rural Electrification Projects. European Journal of Business Management, 2014(1). [9] Phillip.W. Kingori, and Dr. Karanja Ngugi. Determinant of Procurement Performance at Retirement Benefit Authority in Kenya. European Journal of Business Management, 2014(11). [10] Martin I. Kestenbaum, and Ronald L. Straight. Procurement Performance: Measuring Quality, Effectiveness, and Efficiency. Public Productivity & Management Review, 1995(19). [11] Huang Mingjin. On Government Procurement Performance Evaluation [J]. Economic Research Reference, 2009 (5). [12] Peng Xianggang, Xiao Ailin. Analysis on the Innovation of Government Procurement Performance Management Based on Public Value [J]. Journal of Beijing Union University, 2014 (4). [13] Ministry of Finance. China Government Procurement Yearbook 2013 [N]. China Financial and Economic Publishing House, [14] Treasury Department Treasury Department. A summary of national government procurement in 2015 [EB / OL]. [15] Ministry of Finance. China Government Procurement Yearbook [N]. China Financial and Economic Publishing House, [16] Treasury Department Treasury Department. A summary of national government procurement in 2016 [EB / OL]. [17] Treasury Department Treasury Department. A summary of national government procurement in 2014 [EB / OL].