Political risk distribution of Chinese outward foreign direct investment

Abstract

Purpose

The purpose of this paper is to contribute to a better understanding on relations between Chinese Outward Foreign Direct Investment (OFDI) and host country political risk. To contribute to a better understanding of whether traditional wisdom on foreign direct investment (FDI) is sufficient to explain the internationalization of Chinese multinational enterprises, the author collected 15 proxy variables from the PRS Group and Heritage Foundation and applied principal component analysis (PCA) to construct a new political risk index (PRI) that measures multiple facets of political risk for 139 countries.

Design/methodology/approach

Using this new PRI as a criterion, the author investigated changes in the political risk distribution (PRD) of Chinese outward FDI (OFDI) regarding investment destinations, large projects, annual investment outflows and sectorial distributions from 2006–2017.

Findings

The author found that the vast majority of Chinese OFDI during this period is concentrated in moderate- and low-risk countries, even at the sectorial level. This paper also shows that the continuing reform of Chinese OFDI policy and strong government support have led to an unprecedented increase in Chinese OFDI, while the PRD of Chinese OFDI has maintained a gradual decline over the past decade.

Originality/value

This research provides a new measurement that covers multiple facets of political risk.

Citation

GaoYan, Q. (2021), "Political risk distribution of Chinese outward foreign direct investment", International Journal of Emerging Markets, Vol. 16 No. 6, pp. 1202-1227. https://doi.org/10.1108/IJOEM-06-2018-0344

Introduction

In recent years, with the dramatic increase in Chinese outward foreign direct investment (OFDI), the topic of how the political risk in host countries influences the locational choices of Chinese multinational enterprises (MNEs) has drawn the attention of many scholars. Many studies have found that the investment of Chinese MNEs is generally attracted to (Buckley et al., 2007; Li and Liang, 2012; Huang and Wang, 2011; Duanmu, 2012, Quer et al., 2011), deterred by (Cheung and Qian, 2009; Duanmu, 2011; Hurst, 2011; Blomkvist and Drogendijk, 2016) or indifferent to (Kolstad and Wiig, 2012) host country risks. Therefore, some scholars (Rodríguez and Bustillo, 2011; Ramasamy et al., 2012) claim that traditional wisdom on foreign direct investment (FDI) cannot explain the internationalization of Chinese MNEs. However, these claims should be further explored because of their methodological shortcomings. The primary reason for these contradictory results is the lack of a clear articulation of the assessment methodologies of political risk. Another limitation concerns the reliability of data on Chinese OFDI from official sources. My contribution to the literature on the linkages between political risk and Chinese OFDI is twofold. First, this paper presents a new and comprehensive measurement of political risk – the political risk index (PRI) – by using principal component analysis (PCA) to scientifically determine the final weights (FWs) of 15 relevant proxies from reliable data sources. Second, this paper empirically analyzes a less known topic – changes in the distribution of Chinese OFDI in countries with different levels of political risk – over the past decade using China Global Investment Tracker (CGIT) data that clearly show the final destination and ownership of each large-scale investment project undertaken by Chinese MNEs, thus contributing to the debate on whether the relationship between political risk and Chinese OFDI follows conventional wisdom.

The rest of this paper is organized as follows: Section 2 provides a literature review on the concept and measurement of political risk and some empirical works that explore the relations between political risk and Chinese OFDI. Section 3 presents a detailed description of the research methodology. Section 4 empirically analyzes the distribution of Chinese OFDI from 2006 to 2017 using my new PRI as a criterion, and Section 5 concludes the paper.

Literature review

Here, I analyze the concept and measurement of political risk because they provide the theoretical background to this study. I also present conclusions of previous studies on the relationship between Chinese OFDI and political risk.

Political risk and its measurement

Although agreement on the definition of political risk remains contentious, when discussed in the FDI literature, consensus has been reached by associating this concept with the host country's macro environment. According to Alon and Herbert (2009), Lawton et al. (2013 and 2014) and John and Lawton (2017), political risk is a complex and multidimensional phenomenon that may arise from a variety of sources that depend on the stage of institutional and economic development of a host country. It is likely that MNEs will encounter non-violent political risks, such as unfavourable legal rulings and stringent entry requirements, or more severe risks, such as the overthrow of political regimes, wars and expropriations, in host countries with an underdeveloped socio-political and economic environment, in contrast to countries with well-established socio-political and economic systems.

To provide better forecasts for multinational entities, various political risk analysis techniques have been constructed by different companies, organizations and research institutions, with the rating system being the most commonly used technique. Rating systems (Table 1) usually contain political, social and economic models and can be used for scenario analysis to obtain aggregate measures of political risk. These models all share some basic features, namely, a reliance on the judgement of country experts and the subjectivity of the weights assigned to risk factors and indicators. However, common methodological flaws, such as the inclusion of irrelevant variables, arbitrary weightings among relevant proxies and problems of independence, objectivity and transparency, have been widely criticized by scholars.

Review on the relations between political risk and Chinese outward foreign direct investment

With China's ascension to the position of the world's second largest investor, the topic of Chinese OFDI has increasingly drawn many scholars' attention. The more recent literature has explicitly recognized home and host country institutions as essential determinants of inward FDI (Amighini et al., 2014). Following this institutional-based view (IBV), researchers continue to investigate why the economic, institutional and political characteristics of host countries make them attractive to Chinese MNEs (Lattermann et al., 2017), identifying market size (Zhang and Daly, 2011), the exchange rate level (Buckley et al., 2007), natural resource endowments (Kolstad and Wiig, 2012), technology (Alon et al., 2013), distance (Blomkvist and Drogendijk, 2016) and culture (Quer et al., 2011) as major determinants of Chinese OFDI; in this regard, they draw conclusions that are similar to those of research on the internationalization of American, Japanese and Western European MNEs (Jimenez, 2011; Jimenez et al., 2014).

However, in regard to the linkages between political risk and Chinese OFDI, empirical work has often obtained mixed results (Table 2): on the one hand, scholars such as Buckley et al. (2007), Li and Liang (2012), Huang and Wang (2011), Duanmu (2012) and Quer et al. (2011) have reported that high political instability is not a deterrent to Chinese OFDI, but a “catalyst” for Chinese MNEs; on the other hand, Cheung and Qian (2009), Duanmu (2011), Hurst (2011) and Blomkvist and Drogendijk (2016) have found that Chinese MNEs, similar to MNEs from more developed economies, are also deterred by host country political risk; however, other scholars, such as Kolstad and Wiig (2012), have reported the nonexistence of relations between the country risk level and Chinese OFDI and argued that political risk does not necessarily or significantly influence Chinese MNEs' location decisions because presently, host country political risk can be effectively mitigated through either sophisticated strategies or insurance schemes.

The first explanation for the mixed results noted above is the measurement of Chinese OFDI. The vast majority of past studies have used official data sourced from China's Ministry of Commerce (MOFCOM); however, there is a broad consensus among researchers that some problems of underestimation might arise for a number of reasons. First, the collection of MOFCOM data does not follow international standards: rather than gathering information through surveys after the full value of the transaction is credited (when the investment or construction activity has actually started), information is collected during the approval process, therefore causing the problem of underreporting; private firms, whose investments are treated differently depending on the specific regional regulations, are often left out; also excluded are two other situations in which investments do not require formal approval and investments whose approval procedure was not controlled by the MOFCOM (such as financial institutions). Second, MOFCOM data often fail to capture the true origin and final destinations because they provide only aggregate Chinese OFDI in certain recipient countries. Regarding destinations, MOFCOM data are strongly affected by the practice of round-tripping, i.e. the channelling of large investment outflows through tax havens, establishing special purpose entities and reinvesting in China or third countries (Amighini et al., 2014). Round-tripping is the result of a common practice among firms of registering only the first destination of their investments, which results in the overestimation of some transit locations (such as Hong Kong and Macao) concerning the final destinations. Another serious issue concerns the origin of Chinese OFDI. Although the business affairs of firms' owners and principles are physically present in mainland China, many Chinese MNEs have moved their incorporation to tax havens such as Hong Kong, the British Virgin Islands and the Cayman Islands. The methodological result of official FDI statistics often excludes information on these firms and is, therefore, inherently skewed towards the investment patterns and behaviours of companies registered in mainland China. Third, MOFCOM data do not follow the International Standard Industry Classification system but are based on a domestic classification; therefore, the sectorial distribution of Chinese OFDI is inaccurate. For example, “energy” is not a designated sector, but “business and leasing services” is; such an idiosyncratic sector classification often prevents detailed international comparisons.

Apart from the reliability of data on Chinese OFDI, I also find a lack of coherence in variables related to the measure of political risk. The vast majority of studies that consider the impact of political risk on the international location choice of Chinese OFDI often use variables sourced from (1) the PRS Group's International Country Risk Guide (ICRG), (2) the World Bank's Worldwide Governance Indicators (WGI) or (3) the Heritage Foundation's Index of Economic Freedom (IEF). However, their measurement of political risk is sometimes incomplete because scholars usually select only one or several proxy variables, such as rule of law (Huang and Wang, 2011; Kolstad and Wiig, 2012), corruption (Duanmu, 2011; 2012; Kang and Jiang, 2012), overall political stability (Buckley et al., 2007; Cheung and Qian, 2009; Quer et al., 2011; Blomkvist and Drogendijk, 2016) and regulatory quality and protection of property rights (Hurst, 2011; Li and Liang, 2012), from the abovementioned assessment databases, covering only partial aspects of “political risk”. In this way, scholars fail to capture the multiple aspects of political risk.

To fill these gaps, in this study, I intend to do the following: first, I construct a new PRI in an attempt to capture the multiple facets of political risk by introducing 15 relevant proxies. Second, I explore a less known topic – changes in the distribution of Chinese OFDI in countries with different levels of political risk – over the past decade using CGIT data and thus contributing to a debate on whether the relationship between political risk and Chinese OFDI follows conventional wisdom.

Materials and methodology

To construct a comprehensive and objective PRI, I selected a number of sub-indicators from three different data sources and then chose the most relevant proxies using questionnaires collected from Chinese experts and managers with overseas project management experience; I then applied PCA to determine the weights for each proxy. For the measurement of Chinese OFDI, I use CGIT data from the American Enterprise Institute (AEI), while the data on host country political risk are derived from the IEF of the Heritage Foundation, the ICRG of the PRS Group and the WGI by the World Bank.

Measurement of Chinese outward foreign direct investment

I use the CGIT database developed by the AEI and Heritage Foundation as the measurement of Chinese OFDI; this database presents detailed information on Chinese OFDI from 2005 onwards. The database is limited to transactions of US$100m or more (so-called large-scale projects) and consists of 1,406 overseas investments, 1,502 construction projects and 254 troubled transactions for a total of 3,162 large-scale projects. It is one of the most comprehensive, objective and transparent (as it never hides troubled transactions) datasets on Chinese OFDI at the firm level. The CGIT database provides information at the level of the individual deal; the main deal-level information includes (1) the investing company and ownership, (2) the investment value, (3) the sector of specialization of the investing company, (4) the transaction party, (5) the destination of the investment and (6) the investment type (greenfield or non-greenfield).

The advantages of using CGIT data, as opposed to the official database, are that CGIT data are renewed and updated every six months to include changes in earlier years as better information becomes available, while MOFCOM data often ignore performing revisions of annual totals in calculating growth and country shares. Second, the CGIT database clearly provides not only the final destinations but also the origins of Chinese OFDI, which is important because, due to the long-standing Chinese policy, MOFCOM data do not disclose the final destinations of Chinese OFDI after flowing into tax havens; with clear information on the names of the origin investing companies as well information on the transaction party and final destination of investment, the CGIT database, therefore, provides more reliable data. Third, MOFCOM data use an idiosyncratic sector classification; therefore, the sectorial distribution of Chinese OFDI is inaccurate. However, the CGIT database uses conventional categories and frequently updates the list to match new areas of investment, such as tourism and entertainment, facilitating international comparisons. Most importantly, all the deals included in the CGIT database are linked to the firm level (both investor and target companies), thus making it possible to investigate the distribution of investments across sectors, business activities and countries and distinguish by deal type, company and group. Moreover, this database enables the foreign expansion strategies of Chinese firms and groups to be mapped in a more comprehensive way, compared to what could be done previously with non-comparable data on different types of foreign activities. Additionally, at country and regional levels, it allows the location choices to be examined disaggregated by sector and deal type.

Data sources on political risk and the selection of sub-indicators

To measure the multiple facets or multiple dimensions of political risk and to avoid being constrained by one or two widely used variables that capture only one side of “political risk”, I construct a new measurement – the PRI, which covers a more comprehensive range of proxies. I collect data on political risk components from the ICRG, IEF and WGI. There are three advantages of these databases: first, they account for a broad range of indicators that are more relevant to the political environments in host countries. Second, they include a large number of sample countries over a long period. Third, the three databases are already highly regarded among foreign investors when making investment decisions because of the authenticity and objectivity of their information about political institutions and governmental regulations.

To select the most appropriate proxies, I designed a questionnaire containing all sub-indicators that describe the institutional environment of host countries from the three political risk assessments noted above. I disseminated this questionnaire among Chinese scholars and managers with overseas project management experience. The questionnaire was pretested for validity by a panel of experts in the related field (including executives, consultants, scholars and government officers). My questionnaire was distributed among 305 respondents through online links, and I collected 74 valid responses. From a given list containing 33 sub-indicators, the respondents were asked to rate the most relevant proxy variables that influence the operation and profitability of an overseas project based on a five-point rating system, with 5 being the most significant impact.

The results are shown in Appendix 1. Cronbach's alpha of the 33 items is 0.914, indicating that the internal consistency of the times in this scale is assured because, in general, a value of 0.6 is considered adequate for internal consistency (Tung et al., 2012). As the average score of all the items is above 3 (Appendix 2), I ran another t-test (Appendix 3) and found that scholars and managers hold different opinions on the relevance of several sub-indicators of political risk. I deleted all sub-indicators where such disagreements exist and retained only 15 proxy variables; by doing so, to some degree, I avoid the problem of the “irrelevance of the selected proxies”.

Among the remaining 15 proxy variables (see their descriptions in Table 3), my respondents chose 12 variables from the ICRG, while the remaining three were chosen from the IEF. The 15 proxy variables were then categorized into three types of political risk based on their causes. The first group, institutional risk, originates from uncertainties created by policy instability and arbitrary regulation in FDI-related policies; I included government unity (GU), legislative strength (LS), popular support (PS), property rights protection (PRP) and government integrity (GI) in this group. The second group, transfer and expropriation risk, stems from the nationalization or expropriation of foreign assets or breach of contract; contract viability (CV), payment delays (PD), profit repatriation (PR), investment freedom (IF) and exchange rate risk (ERR) belong to this group. The third group, political violence risk, refers to war and other types of political violence that immediately damage foreign assets and discourage the productivity of a host economy in the long term. The remaining proxies, i.e. ethnic tensions (ET), external conflict (EC), internal conflict (IC), military in politics (MP) and religious tensions (RT), are included in this group.

Using Principal Component Analysis (PCA) to Calculate the Final Weight(FW) of Each Proxy

PCA has been widely applied to the construction of essential indicators. By calculating the correlations between variables, PCA is a statistical technique used by scholars to scientifically determine the weights of different variables. Applying PCA to determine the weights of sub-indicators is also a widely accepted method in the literature, especially for topics concerning the construction of new “indices” or “criteria”. Examples include the globalization index (Dreher, 2006), the capital account openness index (Chinn et al., 2008), the regional economic integration index (Chen and Woo, 2010) and currency internationalization prospect indices (Tung et al., 2012). To solve the problem of arbitrary weighting of the selected variables, which is quite common in previous political risk measurement methodologies, I adopt this method to scientifically calculate the FWs of the 15 relevant proxies. I studied 139 economies from 2006–2017 and accumulated a total of 1,390 observations. The linear trend interpolation method was employed to deal with missing data. As the data collected from the ICRG and IEF are marked using different score ranges – for instance, ICRG scores vary between 0 and 12, while IEF scores range from 0–100 – I normalized the data so that all the converted data fell into the range of 0–10. In this way, I could better compare the scores:

(1)C1=10×[(ICRGi−ICRGimin)/(ICRGimax−PRSimin)]C2=10×[(IEFk−IEFkmin)/(IEFkmax−IEFkmin)]

In this equation (equation (1)), C₁ and C₂ represent normalized indicators extracted from the ICRG and IEF, respectively. As previously discussed, the maximum score of the transformed data is 10, while the minimum score is 0. High scores indicate a low level of political risk, whereas low scores indicate a high level of political risk.

As seen in Table 4, the Kaiser–Meyer–Olkin (KMO) value is over 0.6, while the value of Bartlett's test of sphericity was significant at the 1% level. Thus, the suitability of my sub-indictors was confirmed. The Cronbach's alpha scores were higher than the threshold of 0.7, signifying that the chosen 15 proxy variables are highly relevant to the measurement of political risk. Furthermore, the first eigenvalue of the PCA (6.57467) is far greater than the criterion (more than 1.00) of the eigenvalues. However, the explained variance of the first principal component is lower than the standard criterion of 80%, implying that incorporating only the first principal component scores as the proxy weights of the PRI may result in low construct validity (Tung et al., 2012). After calculating the factor loading matrix α_xy in Table 5, I used the following equation (2) to calculate the FW of the political risk proxies:

(2)FWx=∑y=115axy2∑x=115∑y=115axy2

where y is the number of principal components, from PCA1 to PCA15 (y = 1, 2, … 15). x denotes the number of political risk proxy variables (x = 1, 2, 3, …, 15). I made the sum of the 15 political risk proxies equal to unity by using equation (2). This method considers all the primary components and thus extends the total cumulative explained variance to 100%. As presented in Table 6, the FW of each proxy variable can be drawn as follows: ET has the highest weight (9.9%), followed by RT (7.7%), CV (7.6%), LS and IC (7.5%), IF, PD and EC (7.1–7.4%) and PRP, PS, PR, MP, GI, government cohesion and ERR (3.9–6.4%).

Calculation of political risk index and country classification

Using the method discussed in the previous section, I developed the following equation (equation (3)) to calculate the PRI of my 139 sample countries:

(3)PRI=7.6%×CV+6.3%×GU+7.5%×LS+7.2%×PD+4.8%×PS+3.9%×PR+6.4%+ERR+9.9%×ET+7.4%×EC+7.5×IC+5.5%×MP+7.7%×RT+5.2%×PRP+5.9%×GI+7.2%×IF

I use China as an example to demonstrate the calculation of the PRI. In 2017, China had a CA score of 5.6. Using my method, I find that CA accounts for 7.6% of a series of sub-indictors. I multiplied the original CA score by its FW to obtain 0.43. I then summed all 15 weighted scores and obtained China's PRI in 2017, which was 5.42 (Table 7). As previously defined, the PRI ranges from 0–10. A higher PRI indicates lower political risk. Based on the calculated PRI of my samples, I divide the 139 economies into three groups. Countries with a PRI higher than 7.5 were considered low-risk countries; those between 5 and 7.49 were classified as moderate-risk countries; and those below 4.99 were considered high-risk countries.

Results and discussion – political risk distribution of Chinese outward foreign direct investment

In this session, I calculate and rank the PRI of 139 countries and then investigate the political risk distribution (PRD) of Chinese OFDI from 2006 to 2017.

Political risk index of 139 sample countries

I calculated the PRI of the 139 sample countries (Appendix 4) and found that in 2017, 21 countries fell into the low-risk group, 95 were in the moderate-risk group and 22 belonged to the high-risk group. In looking at the trend between 2006 and 2017 (Figure 1a), I found that the number of moderate-risk countries continually increased but stayed within the range of 80–95 countries, stably accounting for 59–68% of the sample. Over the same period (Figure 1b), the number of low-risk countries dropped from its 2007 peak of 42 to only 21 by 2017, showing a decline from 30 to 13%. The number of high-risk countries, however, fluctuated greatly. From 2007–2012, the number rose from 15 to 32 (between 11 and 24%); then, from 2013–2017, it dropped to 22. Although low-risk countries significantly outnumbered high-risk countries from 2006–2011, the proportion of high-risk countries subsequently exceeded that of low-risk countries. Meanwhile, the proportion of moderate-risk countries remained the largest.

Political risk distribution of Chinese outward foreign direct investment

In this section, I calculate the annual PRI of the 139 destination countries covered in the CGIT database from 2006–2017. I first analyze the general trend and then present the distribution of Chinese OFDI in low-, moderate- and high-risk countries.

General trend

According to Figures 2, 3 and 4, I observed that Chinese MNEs do not follow the incremental internationalization pathway suggested by traditional theories such as the Uppsala Model and lifecycle theory (Jiménez, 2010; Jiménez et al., 2014) but expand at a much faster pace. From 2006–2017, Chinese OFDI destinations nearly tripled, increasing from 28 in 2006 to 80 in 2015 and declining to 59 in 2017 (Figure 2), while the annual number of Chinese large-scale FDI projects (those with a single investment of more than US$100m, regardless of construction contractor direct investment) shot up from 49 in 2006 to 408 by 2016, only to fall to 157 in 2017 (Figure 3). I also found that the annual outflows of Chinese OFDI increased six-fold, from US$40.23bn in 2006 to US$261.1bn in 2016, and then dropped to US$132.24bn in 2017 (Figure 4). Regarding industrial distribution, according to Table 8, I found that from 2006–2017, Chinese investment in energy, transport, real estate, raw materials and metals exceeded 70% of its total investment. This finding reflects that Chinese OFDI is mainly concentrated in the field of natural resource development, energy and infrastructure construction.

The decade-long expansion of Chinese MNEs can be divided into two different stages: the “Going Global” phase from 1999–2012 and the “Belt and Road” phase from 2013 onwards. The “Going Out Policy” phase (1998–2013) was marked by clear aims of pushing Chinese domestic enterprises into global business activities to acquire strategic resources and expand into foreign markets. During this period, the “home-country network ties” theory (Lattemann et al., 2017) helps explain the internal expansion of Chinese MNEs. At this stage, the Chinese government encouraged large-scale state-owned enterprises (SOEs) and powerful privately owned enterprises to acquire strategic resources, expand into foreign markets and invest in key “sensitive” industries defined by the 12^th and 13^th Five-Year Plans, such as cutting-edge technology, renewable energy, energy saving and environmental protection; in this regard, “state involvement and the government's industrial policy are central to firms' international expansion”. The Chinese government steadfastly supported Chinese MNEs' overseas activities through political backing, subsidies, preferential tax concessions, the reformation and relaxation of the regulatory process and the easing of foreign exchange controls.

When President Xi Jinping announced the “One Belt, One Road” (OBOR) initiative in 2013, it was believed to be an updated version of the “Going Global Strategy”, with a clear aim of better integrating the Chinese economy with the economies of its neighbouring countries in Central Asia, Southeast Asia and South Asia as well as those of Eastern Europe and the Baltics through infrastructure and production capacity cooperation. During the OBOR phase (from 2013 onwards), the expansion of Chinese OFDI into OBOR countries can be better explained by the theory of the “level of interaction and relations” (Jiménez, 2011), such as economic diplomacy initiatives, the general level of economic trade and investment flows and strong political ties between China and the target country. Through 2018, more than 86 sovereign states and international organizations have signed OBOR cooperation agreements with China, and the majority of these countries have favoured the “Beijing consensus” and have been more likely to let “both formal and informal institutions develop under the role of the state” (Lattemann et al., 2017).

However, in 2017, the Chinese government started to impose restrictions on the overseas activities of Chinese companies, curbing what it deems “irrational buying sprees” by Chinese MNEs, especially in the real estate and entertainment sectors, because the Chinese government considered large-scale currency outflows in the form of unrelated diversification acquisitions to be a significant threat to its ability to govern. Including the management of China's official exchange rate, it is clear that this form of institutional arbitrage took place in significant volumes; the effects of these restrictions explain why the number of countries with large Chinese projects shrinks in 2017, but the lasting effects of this legislation (adopted in late 2017) may not appear for some time.

Distribution of Chinese outward foreign direct investment in low- and moderate-risk countries

As seen from my figures, from 2006–2017, over 11 years, moderate-risk countries continue to be Chinese MNEs' priority targeting investment economies, comprising 45–56% of the annual investment destinations and hosting 37–53% of the annual large-scale overseas projects. Regarding annual capital flows, the percentage of annual Chinese OFDI in moderate-risk countries fluctuates between 28 and 66%. Despite drastic changes, moderate-risk countries continued to absorb the majority (over 45%) of China's annual OFDI flows between 2006 and 2015. Regarding sectorial distribution, I found that moderate-risk countries host more than half of China's aggregate investment in the energy, transportation, metals, utilities and chemical industries. The real estate sector is somewhat unique, with China's total investment being evenly distributed between moderate- and low-risk countries. During the same period, low-risk countries accounted for 11–25% of Chinese MNEs' annual investment destinations, hosted 18–42% of China's annual overseas large-scale projects and absorbed 14–64% of China's annual OFDI outflows. At the sector level, more than 56–65% of China's total investment in agriculture, science and technology and finance was located in low-risk countries; for industries such as tourism and entertainment, this number exceeds 70%.

The fact that the majority of Chinese OFDI is located in low- and moderate-risk countries proves that the level of political risk in host countries is not ignored by Chinese MNEs, thus verifying that traditional FDI theories can sufficiently explain the locational choices of Chinese MNEs. However, other reasons may also help explain this phenomenon. On the one hand, troubled transactions have taught Chinese MNEs to thoroughly evaluate the possible political risk in host countries and to take more cautious procedures and steps before making OFDI decisions. On the other hand, through effective government guidance, Chinese MNEs avoid making wrong decisions. The first type of guidance is “information guidance”. By providing Chinese enterprises with detailed, specific and up-to-date information on the targeting host countries, the Chinese government helped reduce the PRD of Chinese OFDI. Taking the country (region) guidance for OFDI and cooperation issued annually by the MOFCOM as an example, when first published in 2009, it contained warnings of the possible adverse effects on Chinese OFDI caused by the global financial crisis. Subsequently, however, comprehensive information (including important political, social and economic events in each country that may impact the political risk of investments) that explained the political risk profile of individual countries was provided. Currently, investors can even find possible mitigations for dealing with certain types of political risks. The guidance itself is now regarded not only as an academic analysis but also as an investment roadmap for Chinese entrepreneurs with OFDI ambitions. Chinese financial institutions, especially the China Export and Credit Insurance Corporation, also contribute by providing the national risk analysis report of 60 major economies and insurance mechanisms that help Chinese enterprises effectively mitigate political risk.

The second type of guidance is “policy guidance”. Through explicit policies, the Chinese government has helped improve the PRD of Chinese OFDI. The Chinese government has made use of the fact that Chinese OFDI is mostly carried out by SOEs and is, therefore, in a better position to implement its OFDI guidance. For instance, from 2005, the annually updated blacklist compiled by the State-owned Assets Supervision and Administration Commission (SASAC) recorded all historically bad investment decisions made by Chinese SOEs. It is now used not only as a warning for SOEs to avoid political risks in their OFDI activities, but also as an important performance check for senior managers in those SOEs. To further strengthen the capacity of its enterprises to manage major emergencies overseas, the SASAC has also issued other guidelines entitled the “Overseas Property Management Interim Measures” and the “Overseas Assets Supervision and Administration Interim Measures”. Another example is the “Opinions on State-owned Enterprises to Promote the Implementation of Three Crucial Decisions and a Big Decision-making System”, a guideline jointly issued by the Chinese Communist Party Central Office and the Chinese State Council in 2010; this guideline made it very clear that to prevent losses, any OFDI decisions ought to be made based on consensus among senior managers. The MOFCOM issued another policy, the “Security Risk Early Warning and Information Notification System for Chinese Enterprise OFDI and Cooperation”, which clearly defined the risks involved in OFDI and required Chinese companies' compliance.

Distribution of Chinese outward foreign direct investment in high-risk countries

From 2006–2017, high-risk countries comprised 15–31% of Chinese MNEs' annual investment destinations and absorbed 7–32% of Chinese annual OFDI outflows. Additionally, approximately 15–30% of China's annual overseas large-scale projects were located in high-risk countries. However, in the energy, transportation and utility industries, high-risk countries hosted approximately 30% of the total Chinese investment.

According to traditional assumptions, high levels of political risk have been understood as a threat to MNEs; however, many studies have found that Chinese MNEs prefer to invest in high-risk countries. According to the results, the majority (over 90%) of large-scale projects in high-risk countries are undertaken by SOEs. This result can be explained by the institutional factors in China, especially the unique “state ownership advantage” brought by the unique “government–business” relationship between SOEs and the central government. This relationship has made SOEs less dependent on their own and even on other Chinese firms' prior host country experience because under this “government–business” relationship (Quer et al., 2018), SOEs come under the direct supervision of the SASAC, and the government provides them with political backing while SOEs are put in place to implement the government's “Going Out” policy and OBOR initiative.

Second, the results also show that the OFDI undertaken by SOEs has achieved satisfactory performance (only 32% are classified as troubled transactions), proving that institutional factors are not the sole reasons for the reduced PRD of Chinese OFDI. As Holburn and Zelner (2010) show, because there are weaker institutional constraints on policymakers and greater redistributive pressures associated with political rent-seeking in China's business environments, all companies, regardless of their ownership, have developed strong political capabilities through organizational learning and cognitive imprinting. Thus, they know how to operate in the challenging institutional environment – comprising a high level of direct state intervention, opaque corporate governance, unpredictable and burdensome regulations, cumbersome bureaucracy and discontinuity in government policies – that characterizes the Chinese business environment. As a result, they are more capable of dealing with burdensome regulations and navigating opaque political constraints (Buckley et al., 2007). For many Chinese firms, their familiarity with the more difficult institutional conditions of developing countries and their expertise in managing such environments have helped them develop strong political capabilities, such as the negotiation of entry conditions, lobbying, litigation, campaign contributions and coalition formation, leading to preferential conditions, reduced environmental uncertainty, reduced transaction costs and increased long-term sustainability for the firm, which reduce the deterrent effect of political risk on their foreign entry decisions. Sometimes, they even obtain a better competitive advantage over their western competitors, as the latter are not used to the absence of a well-established infrastructure and a well-developed contracting and intellectual property rights regime in high-risk countries (Cueruo-Cazura and Genc, 2008).

To summarize, the following conclusions can be drawn: first, by applying this PRI, I found that between 2006 and 2017, low- and moderate-risk countries remain Chinese companies' predominant investment destinations; however, the majority of Chinese OFDI is undertaken by Chinese SOEs. Second, I observed that until 2017, the value of China's overseas investment and construction combined was approaching US$1,6325tn. During the same period, Chinese MNEs invested in a total of 2,276 large-scale projects worldwide. Of these, 677 large-scale projects were located in low-risk countries, another 1,066 took place in moderate-riskcountries and 533 occurred in high-risk countries. In terms of aggregate investment, low-risk countries were found to host 33% of the total stock of Chinese OFDI, while moderate-risk countries accounted for 45%, and the remaining 22% of Chinese OFDI was in high-risk countries. The majority of Chinese OFDI is, therefore, directed towards moderate- and low-risk countries. Third, during the same period, low- and moderate-risk countries account for more than 70% of the total Chinese OFDI in 12 different industries. Only in industries such as energy, transportation and utilities did high-risk countries host approximately 30% of the total Chinese OFDI. Among the remaining nine industries, high-risk countries attracted only a small proportion of China's total OFDI.

Conclusion

By increasing its foreign investment linkages with other countries, China has integrated rapidly with the world economy. It is found that through continuing international economic policy reform, effective government guidance and the continuing development of political capabilities within Chinese MNEs, Chinese OFDI has witnessed a substantial increase in the past ten years, while the political risk of Chinese OFDI has declined rapidly.

The contribution of this paper is twofold. First, this paper presents a new measurement of political risk – the PRI – by using PCA to scientifically decide the FWs of 15 relevant proxies from reliable data sources. Second, this paper empirically analyzes a less known topic – changes in the distribution of Chinese OFDI in countries with different levels of political – over the past decade using CGIT data that clearly show the final destination and ownership of each large-scale investment project undertaken by Chinese MNEs, resulting in certain findings that contradict many scholars' research on the “political risk aversion” characteristic of Chinese MNEs during their internationalization: both low- and moderate-risk countries remain Chinese MNEs' predominant investment destinations in terms of the number of large-scale projects, annual investment outflows and sectorial distributions. My conclusion proves that the importance of the political risk level in host countries has never been neglected by Chinese MNEs when making investment decisions because troubled transactions have taught them to take more cautious procedures and steps in their OFDI; at the same time, governmental efforts, such as effective guidance, consultation and real-time investment information, have prevented Chinese enterprises from investing in troubled projects and high-risk-level countries. Most importantly, the political capacities developed by Chinese MNEs in the Chinese business environment have helped them overcome many difficulties in their internationalization.

Regarding limitations and further research, clearly, this research is limited by the lack of comparisons between the PRD of Chinese OFDI and that of other countries due to the unavailability of data on their OFDI distribution (to date, I have not found any other databases that provide detailed information on Chinese OFDI similar to the CGIT database provided by the AEI or EMENDATA). However, given more data, future research could be conducted to explore the similarities and differences between the PRD of Chinese OFDI and that of other countries. Comparisons could be made between developed economies, such as those of the USA, Japan and Western Europe, and between emerging economies, such as those of Russia, Brazil, India and South Africa, to better understand the locational choices of Chinese enterprises during international expansion.

Figures

Figure 1a

Political risk distribution of 139 countries (numbers), 2006–2017

Figure 1b

Political risk distribution of 139 countries (percentage), 2006–2017

Figure 2

PRD of Chinese OFDI destinations, 2006–2017

Figure 3

PRD of Chinese OFDI large-scale projects, 2006–2017

Figure 4

PRD of annual Chinese OFDI outflows (US$m), 2006–2017

Table 1

Political risk assessment models by different institutions

Institution	Categories and weight	Number of countries covered	Definition of political risk	Political risk components	Source of data
Thomson Reuters country risk ranking	Political risk (50%)	240	NA	Type of governance Civil liberties and political rights Media freedom Government effectiveness Regulatory quality Rule of law Political terror scale Armed conflict Human rights Failed states index	Survey respondents and experts judgement
PRS group International Country Risk Rating (ICRG)	Political risk rating (50%)	141	“A means of assessing the political stability of the countries covered by ICRG on a comparable basis”	Government stability Socioeconomic conditions Investment profile Internal conflict External conflict Corruption Military in politics Religious tensions Law and order Ethnic tensions Democratic accountability Bureaucratic quality	Expert judgement
EIU country risk service	Political risk (20%)	131	A range of political factors relating to political stability and effectiveness that could affect a country's ability and/or commitment to service its debt obligations and/or cause turbulence in the foreign-exchange market	External conflict Governability Electoral cycle Orderly transfers Event risk Sovereignty risk Institutional effectiveness Corruption Corruption in the banking sector Commitment to pay	Expert judgement
China export and credit insurance company – SINO RATING	Political risk (25%)	192	“Examination of the stability of political institution and the diplomatic relations with other countries”	Political institution Political stability International relations	Expert judgement

Table 2

The relationship between Chinese OFDI and political risk

Author	Time span	Number of sample countries	Measurement of political risk	Measurement of Chinese OFDI	Direction of influence
Buckley et al. (2007)	1984–2001	49	Host country's annual political risk rating	MOFCOM data	+
Cheung and Qian (2009)	1991–2005	46	Host country's annual political risk rating	MOFCOM data	−
Duanmu (2011)	1981–2005	74	Corruption and rule of law	MOFCOM data	−
Hurst (2011)	2003–2008	154	Property rights	MOFCOM data	−
Huang and Wang (2011)	2003–2009	63	Rule of law	MOFCOM data	+
Kang and Jiang (2012)	1995–2007	104	The composite index of 6 elements	MOFCOM data	−
Duanmu (2012)	1999–2008	47	Composite index of 5 element	MOFCOM data	+
Kolstad and Wiig (2012)	2003–2006	142	Rule of law	MOFCOM data	NA
Li and Liang (2012)	2003–2005	95	Property rights	MOFCOM data	+
Quer et al. (2011)	2005–2009	52	Host country's annual political risk rating	MOFCOM data	+
Blomkvist and Drogendijk (2016)	2003–2012	39	Host country's annual political risk rating	MOFCOM data	−

Table 3

Variable description and data source of PRI

Category	Variable name	Description	Source
Transfer and expropriation risk	CV	Risk of unilateral contract modification or cancellation and outright expropriation of foreign-owned assets	ICRG
	PD	Risk associated with receiving and exporting payments from the host country due to political impediments	ICRG
	PR	Risk associated with the difficulty of transferring profits out of the host country due to political impediments	ICRG
	IF	Degree of restrictions as well as rules imposed on foreign investment in a host country due to political reasons	IEF
	ERR	Annual percentage change in the exchange rate of the national currency against the US$ due to political reasons	ICRG
Institutional risk	GU	Extent to which the cabinet coalesces around the government's general policy goals	ICRG
	LS	Extent to which the government can realize its policy programme through the legislative arm of government	ICRG
	PS	Level of support for the government and/or its leader, based on credible opinion polls	ICRG
	PRP	Degree to which a host country's laws protect private property rights	IEF
	GI	Systemic corruption of host country government institutions and decision-making	IEF
Political violence risk	ET	Degree of tension attributable to racial, national or language divisions	ICRG
	EC	Risk to the incumbent government from the foreign action	ICRG
	IC	Political violence in the country and its impact on governance	ICRG
	MP	Military's involvement in politics	ICRG
	RT	Religious tensions	ICRG

Table 4

PCA fit test results

Item type	Statistics	Criteria
Kaiser–Meyer–Olkin appropriateness sampling test	0.875	>0.6
Bartlett test of sphericity	17,662.259***	p < 0.05
Cronbach's α internal consistency reliability	0.8908	>0.7
The first eigenvalue of PCA	6.57467	>1
Construct validity (the first principal component total explained variance)	43.83%	>80%

Source(s): Calculated by the author

Note(s): *p < 0.1; *p < 0.05; ***p < 0.01

Table 5

Eigenvalue, cumulative and factor loading matrix

	y
x	PC1	PC2	PC3	PC4	PC5	PC6	PC7	PC8	PC9	PC10	PC11	PC12	PC13	PC14	PC15
CV	0.32	0.02	−0.24	−0.01	0.12	−0.14	0.03	−0.39	0.04	−0.34	0.32	−0.16	0.63	0.09	−0.07
GU	0.15	0.53	−0.06	−0.18	0.11	0.04	−0.24	−0.06	0.19	0.12	−0.46	−0.57	−0.01	−0.03	−0.00
LS	0.06	0.56	−0.04	−0.21	0.23	−0.04	−0.33	0.23	0.02	0.15	0.37	0.50	0.04	0.09	0.03
PD	0.33	−0.06	−0.23	0.00	0.09	−0.23	0.16	−0.20	−0.11	0.16	−0.26	0.18	−0.28	0.70	0.06
PS	0.11	0.43	−0.18	−0.09	−0.46	0.37	0.61	0.10	−0.14	−0.04	0.06	0.04	0.02	0.00	−0.01
PR	0.33	0.01	−0.28	−0.01	0.08	−0.12	0.06	−0.40	0.03	0.09	−0.01	0.22	−0.37	−0.65	−0.02
ERR	0.11	0.18	−0.10	0.90	0.20	0.25	−0.08	0.08	−0.11	0.07	0.00	−0.01	0.01	0.00	−0.02
ET	0.19	0.10	0.48	−0.00	0.53	−0.22	0.56	0.19	0.09	−0.05	−0.06	−0.01	0.05	−0.11	0.07
EC	0.26	0.10	0.17	0.23	−0.53	−0.49	−0.11	0.20	0.28	−0.10	−0.28	0.24	0.19	−0.08	−0.05
IC	0.30	0.11	0.33	0.03	−0.18	−0.09	−0.19	−0.00	−0.23	−0.40	0.38	−0.30	−0.49	0.08	0.12
MP	0.32	−0.15	0.13	−0.06	−0.14	−0.12	−0.05	0.12	−0.44	0.68	0.17	−0.22	0.24	−0.11	−0.01
RT	0.20	−0.00	0.57	−0.05	−0.07	0.49	−0.14	−0.47	0.03	0.06	−0.17	0.29	0.13	0.06	−0.02
PRP	0.32	−0.23	−0.18	−0.11	0.07	0.25	−0.13	0.31	−0.05	−0.19	−0.20	0.08	0.13	−0.10	0.71
GI	0.33	−0.18	−0.09	−0.16	0.15	0.20	−0.11	0.37	−0.21	−0.26	−0.20	0.08	−0.03	−0.04	−0.67
IF	0.29	−0.22	−0.04	0.01	−0.05	0.24	0.05	0.17	0.73	0.24	0.34	−0.14	−0.14	0.11	−0.08
Eigenvalue	6.57	2.24	1.35	0.93	0.77	0.58	0.52	0.47	0.38	0.28	0.27	0.25	0.19	0.13	0.07
Cumulative	43.8%	58.8%	67.8%	73.9%	79.0%	82.9%	86.4%	89.5%	92.1%	93.9%	95.8%	97.4%	98.7%	99.6%	100.0%

Table 6

FW of each proxy variable

CV	GU	LS	PD	PS	PR	ERR	ET	EC	IC	MP	RT	PRP	GI	IF
7.6%	6.3%	7.5%	7.2%	4.8%	3.9%	6.4%	9.9%	7.4%	7.5%	5.5%	7.7%	5.2%	5.9%	7.2%

Table 7

China's PRI, 2017

Variable	Original score (A)	FW (B)	Weighted score (A)*(B)
CV	5.6	7.6%	0.4256
PD	7.86	7.2%	0.56592
PR	5.28	3.9%	0.20592
IF	5.36	7.2%	0.38592
ERR	6.06	6.4%	0.38784
GS	5.94	6.3%	0.37422
LS	9.82	7.5%	0.7365
PS	5	4.8%	0.24
PRP	4.55	5.9%	0.26845
GI	5.09	5.9%	0.30031
ET	5	9.9%	0.495
EC	6	7.4%	0.444
IC	4.97	7.5%	0.37275
MP	3.98	5.5%	0.2189
PRI	NA	100%	5.42133

Source(s): Calculated by the author

Table 8

PRD of Chinese OFDI at sectorial level

Sector	Agriculture	Chemicals	Energy	Entertainment	Finance	Metals	Real estate	Technology	Tourism	Transport	Utilities	Other	Total
Quota/US$m	91,310	18,980	578,220	37,590	66,560	146,680	156,070	74,800	39,520	275,820	21,760	38,100	1,545,410
High-risk countries	9%	8%	32%	4%	4%	25%	17%	19%	9%	26%	33%	10%	22%
Moderate-risk countries	26%	69%	51%	23%	41%	50%	44%	37%	28%	51%	50%	32%	45%
Low-risk countries	65%	23%	17%	73%	55%	25%	38%	44%	64%	23%	17%	59%	33%
Percentage in total investment quota	5.9%	1.2%	37.4%	2.4%	4.3%	9.5%	10.1%	4.8%	2.6%	17.8%	1.4%	2.5%	100%
Large-scale projects	99	29	657	52	85	204	361	141	58	450	58	82	2,276

Source(s): Calculated by the author

Table A1

Reliability analysis

	N	%
Number of valid questionnaires	74	100.0
Total number of questionnaires	74	100.0

	Cronbach's alpha	Number of sub-indicators
	0.914	33

Table A2

Results of questionnaire

Sources	Sub-indicator	Mean
IEF	Protection of property rights	3.333
	Freedom from corruption	3.748
	Fiscal freedom	3.670
	Government spending	3.415
	Business freedom	3.874
	Labour freedom	3.716
	Monetary freedom	3.881
	Trade freedom	3.615
	IF	4.556
	GI	3.148
	Financial freedom	3.911
WGI	Voice and accountability	4.074
	Political stability and absence of violence	3.748
	Government effectiveness	3.741
	Regulatory quality	3.200
	Rule of law	3.919
	Control of corruption	3.885
PRG	GU	3.489
	LS	3.459
	PS	4.104
	Unemployment	3.970
	Consumer confidence	3.407
	Poverty	3.933
	CV/expropriation	3.333
	PR	3.470
	PD	4.037
	IC	3.830
	EC	3.667
	Corruption	3.481
	MP	3.926
	RT	3.993
	Law and order	3.846
	ET	4.678
	Democratic accountability	3.425

Table A3

Independent samples test

Hypothesis		Levene's test for equality of variations		T-test for equality of means
		F	Significance	t	Significance (two-tailed)	Mean difference
PRP	Ho: Equal variations	0.792	0.375	1.903	0.059	0.36331
	H1: Unequal variations			1.885	0.062	0.36331
GU	Ho: Equal variations	1.605	0.207	−0.733	0.689	−0.07475
	H1: Unequal variations			−0.705	0.686	−0.07475
IF	Ho: Equal variations	0.305	0.582	1.565	0.120	0.19708
	H1: Unequal variations			1.508	0.135	0.19708
GI	Ho: Equal variations	0.005	0.946	0.174	0.862	0.02492
	H1: Unequal variations			0.168	0.867	0.02492
LS	Ho: Equal variations	6.734	0.011	0.419	0.676	0.07546
	H1: Unequal variations			0.440	0.661	0.07546
PS	Ho: Equal variations	1.010	0.317	−1.812	0.072	−0.30850
	H1: Unequal variations			−1.814	0.073	−0.30850
ERR	Ho: Equal variations	0.108	0.744	0.771	0.442	0.12553
	H1: Unequal variations			0.790	0.431	0.12553
CV/expropriation	Ho: Equal variations	0.084	0.773	1.427	0.156	0.20313
	H1:Unequal variations			1.398	0.165	0.20313
PR	Ho: Equal variations	2.895	0.091	1.089	0.278	0.18272
	H1:Unequal variations			1.125	0.263	0.18272
PD	Ho: Equal variations	1.086	0.299	0.581	0.562	0.10252
	H1:Unequal variations			0.563	0.575	0.10252
IC	Ho: Equal variations	0.369	0.545	−0.609	0.543	−0.09635
	H1:Unequal variations			−0.607	0.545	−0.09635
EC	Ho: Equal variations	0.269	0.605	0.660	0.511	0.10014
	H1:Unequal variations			0.666	0.507	0.10014
MP	Ho: Equal variations	6.249	0.014	−1.061	0.291	−0.15876
	H1:Unequal variations			−0.987	0.327	−0.15876
RT	Ho: Equal variations	0.141	0.708	−0.490	0.625	−0.08021
	H1:Unequal variations			−0.486	0.628	−0.08021
ET	Ho: Equal variations	0.088	0.767	−0.734	0.464	−0.12696
	H1:Unequal variations			−0.727	0.469	−0.12696

Table A4

PRI for 139 countries in 2017

Albania	6.664096069
Algeria	4.856290128
Angola	4.875340683
Argentina	6.593757495
Armenia	6.385301118
Australia	7.62246559
Austria	7.394287632
Azerbaijan	5.532944926
Bahamas	7.482189179
Bahrain	6.027351609
Bangladesh	4.989085593
Belarus	5.810579861
Belgium	6.982108641
Bolivia	5.009044416
Botswana	7.323568767
Brazil	5.887262804
Brunei	7.182322426
Bulgaria	6.828628814
Burkina Faso	5.479170693
Cameroon	5.280645392
Canada	8.126311083
Chile	6.997891085
China	5.436418162
Colombia	6.49283082
Congo	5.745268635
Democratic Republic of the Congo	3.494217694
Costa Rica	6.813338391
Ivory Coast	4.971041601
Croatia	6.346077376
Cuba	6.010791107
Cyprus	6.509496481
Czech Republic	7.302913682
Denmark	6.945502217
Dominican Republic	7.372073685
Ecuador	5.095890616
Egypt	5.119579899
El Salvador	6.360532016
Estonia	7.230238562
Ethiopia	4.519090311
Finland	8.155162342
France	6.492600376
Gabon	5.73725031
Gambia	6.230089775
Germany	7.599740407
Ghana	6.169000508
Greece	6.504624737
Guatemala	6.13512804
Guinea	4.486387014
Guinea-Bissau	4.708036237
Guyana	5.950427816
Haiti	5.038475288
Honduras	6.300650347
Hong Kong	8.080549758
Hungary	7.364535873
Iceland	7.74609493
India	5.146929
Indonesia	5.292601781
Iran	5.008832241
Iraq	3.647974502
Ireland	8.03720024
Israel	5.564015809
Italy	7.15398811
Jamaica	7.530451884
Japan	8.332893727
Jordan	6.070199525
Kazakhstan	6.149485065
Kenya	5.574296671
North Korea	5.202430317
South Korea	7.377709942
Kuwait	5.994501688
Latvia	6.925527869
Lebanon	5.186579187
Liberia	4.83149528
Libya	4.708267851
Lithuania	7.014175436
Luxembourg	8.075384342
Madagascar	5.692151255
Malawi	4.9979777
Malaysia	6.144732675
Mali	5.306700347
Malta	7.758266271
Mexico	5.655933207
Moldova	5.540087746
Mongolia	6.482898523
Morocco	6.21429968
Mozambique	5.595319075
Myanmar	4.760022555
Namibia	7.220651426
The Netherlands	7.714877782
New Zealand	7.949435757
Nicaragua	6.44273066
Niger	5.004981168
Nigeria	3.68514984
Norway	8.024324704
Oman	7.308828486
Pakistan	4.179007148
Panama	6.985277839
Papua New Guinea	5.177690578
Paraguay	6.248916739
Peru	6.181778032
Philippines	5.908666343
Poland	7.466284478
Portugal	7.643561294
Qatar	7.393825217
Romania	6.360439547
Russian Federation	5.64455005
Saudi Arabia	6.2355485
Senegal	5.485315303
Serbia	5.825000236
Sierra Leone	5.444772163
Singapore	8.601383809
Slovakia	6.634885264
Slovenia	6.358471199
Somalia	3.054444516
South Africa	6.036175634
Spain	6.797889085
Sri Lanka	4.838511719
Sudan	3.378878395
Suriname	5.814378381
Sweden	8.116488002
Switzerland	8.321517642
Syria	3.803571127
Tanzania	5.426128137
Thailand	5.373564074
Togo	5.611604748
Trinidad and Tobago	6.244860654
Tunisia	5.665589533
Turkey	4.947277301
UAE	7.642741139
Uganda	5.137744507
Ukraine	5.598736672
Britain	7.685671253
USA	7.980181388
Uruguay	7.289377031
Venezuela	3.95059823
Vietnam	5.981801271
Yemen	3.890058749
Zambia	6.345602263
Zimbabwe	4.680232753