How free trade agreement affects the success of China's Belt and Road infrastructure projects
Abstract
After China launched the Belt and Road Initiative in 2013, the Chinese government has invested over 1200 infrastructure projects in countries along the Belt and Road, and nearly 10% of the projects have failed or been delayed. China has also concluded free trade agreements with many countries along the Belt and Road to reduce trade and investment barriers to promote economic and trade cooperation. This paper analyzes whether the signing of the free trade agreement (FTA) is conducive to promoting China's infrastructure investment projects’ success in the countries along the Belt and Road. We conclude that the signing of the FTA agreement is conducive to reducing the failure rate of China's infrastructure investment in the countries along the Belt and Road. For deeper FTAs, the effect is more pronounced. This paper further explores the investment provision of FTA and concludes that investment provision can improve the success rate of investment, especially if it includes provision of national treatment in the pre-establishment and the dispute settlement mechanism.
1 INTRODUCTION
After China launched the Belt and Road Initiative in 2013, the economic and trade cooperation between related countries along the Belt and Road accelerated and upgraded.1 Many railways, ports, underground pipelines, hospitals, and other major projects have been launched, benefiting related countries and regions. Benefits from infrastructure investments, such as relief in traffic pressure, improvement in energy supply, and information development, are instrumental to economic development of the host country. However, investment in infrastructure projects is characterized by a long cycle and a large demand for funds. The projects are subject to many interfering factors such as politics, environmental protection, labor relations, and the potential risks they face are relatively high (Lim et al., 2021; Shaikh et al., 2016). By the end of 2021, the Chinese government-financed infrastructure projects were over 1,200 in countries along the Belt and Road. However, nearly 10% of the projects have failed or been suspended. Therefore, it is crucial to analyze the impact of FTA on the success or failure of such infrastructure projects in these countries.
While promoting the “Belt and Road Initiative,” China has also concluded FTAs with many countries along the Belt and Road to reduce trade and investment barriers to promote economic and trade cooperation. By the end of 2021, China had signed free trade agreements (FTA) with 23 countries along the Belt and Road. With the increase in the number of FTA agreements, it is necessary to systematically evaluate the effects of FTAs signed by China. Unlike the existing research literature analyzing the impact of FTA on the economy, trade, and investment flow, this paper will focus on analyzing whether the FTA concluded by China effectively promote infrastructure projects investment in related countries along the Belt and Road. The results have implications for future FTA content design and overseas investment strategies.
The related literature mainly focuses on analyzing the effect of FTA on foreign investment flows. Yeyati et al. (2003) used the data on bilateral outward FDI stocks from the OECD International Direct Investment Statistics from 1982 to 1999 and concluded that the FTA between the home country and the host country increased bilateral OFDI. Moreover, Baltagi et al. (2008) analyzed the effect of the Europe Agreements on bilateral FDI within Europe; the authors found a significant influence on FDI. However, the effect depended on the geographical location and economic proximity across the host markets. Additionally, Ismail, Smith, and Kugler (2009) used the investment gravity model to study the attractiveness of the establishment of ASEAN (Association of Southeast Asian Nations) to the FDI of countries inside and outside the alliance. The results showed that ASEAN countries in the region were more inclined to invest in new member states, and the results from extra-regional showed that the EU increased its investment in ASEAN countries. Medvedev (2012) used a comprehensive database of PTAs and found that PTAs between developing countries can better increase their proximity and have a positive change in net FDI inflows. Further, Henk and Hugo (2020) found that FTAs average has a direct positive effect on bilateral FDI of around 30%, and deeper FTAS have a larger effect on the FDI flows.
Another part of the literature focuses on the influence channel of FTA on investment. One channel linking FTA with investment is that FTA will bring about trade creation and trade transfer effects, leading to the redistribution of investment resources. Therefore, the impact of FTA on investment between members depends on whether the reduction in trade costs is a substitute or complement to investment. Kindleberger (1996) presented the theory of “investment creation” and “investment transfer” based on “trade creation” and “trade transfer,” and for the first time, tried to combine trade liberalization with investment. Horst (1972) stated that high tariff rates on imported goods induced FDI inflows into Canada, while FTA reduces the tariff and promotes investment. Lakatos and Walmsley (2012) researched the effect of ASEAN–China free trade agreement on investment and provided clear evidence of investment diversion effects in regions that are not signatories to the FTA. Tintelnot (2017) developed a quantifiable multi-country general equilibrium model to analyze the effect of FTA on the multinationals’ decisions and provided evidence for the complementarity between investment and trade.
Some researchers focused on FTA provisions and analyzed whether these provisions affect investment. Markusen and Maskus (2002) concluded that FTAs could reduce the investment barrier, which benefits the exchanging of various inputs and investment efficiency, therefore stimulating more investment. Researchers also studied the detailed protocol in the frame of FTA, including service, investment, and liberalization. Lesher and Miroudot (2006) constructed an investment provision index according to RTA protocols and found that substantive investment rules and provisions can liberalize other parts of the economy and positively promote trade and investment flows. Further, Berger et al. (2013) focused on the investment agreement in the framework of FTA and found that FTA provides free investment rules that can significantly increase the flow of investment. Additionally, Büthe and Milner (2014) analyzed the investment data in developing countries. The authors concluded that preferential trade agreements (PTAs) provide mechanisms for committing to foreign investors about how their assets are handled and are more credible than domestic policies, thus promoting investor confidence and increasing investment. Osnago et al. (2015) studied how deep trade agreements affect foreign investments. The authors concluded that deeper FTA, including standard, intellectual property rights, and investment protection, can reduce contractual uncertainty hence promote investment. Desbordes (2016) believes that establishing a dispute mechanism between investors and the host country in bilateral investment treaty can protect foreign investors from discrimination and significantly promote foreign direct investment. Furthermore, Laget et al. (2021) concluded that protocols related to investment liberalization and protection have a significant positive influence on FDI flow in service-related activities.
Yet, previous studies rarely focused on the influence of FTA on specific investment projects. To the best of our knowledge, this is the first study to investigate the effects of FTA on the success or failure of infrastructure projects. We focus on the implementation of infrastructure projects in countries along the Belt and Road, and analyze whether FTA can promote the success of the investment.
The rest of the paper is organized as follows. Section 2 introduces the empirical model and data employed; Section 3 reports empirical results, including baseline results, robustness checks, and research on FTA heterogeneity. Section 4 further explores the investment provisions included in FTA and analyzes whether these provisions can reduce the failure rate of projects, while Section 5 concludes.
2 METHOD AND DATA
The main explanatory variable is FTA. FTA indicates whether China has a free trade agreement in force with the host country at time t; if is in force at time t, it is defined as 1; otherwise, it is 0. To date, agreements signed between China and the countries along the Belt and Road involved the ASEAN countries, Pakistan, Chile, Peru, Costa Rica, Georgia, and Mauritius. The detailed information is shown in Table 1.
| FTA | Type | Time in force |
|---|---|---|
| China-Aseanb | Multilateral regional free trade agreement | 2005/07 |
| China-Chile | Bilateral regional free trade agreement | 2006/10 |
| China-Pakistan | Bilateral regional free trade agreement | 2009/10 |
| China-Peru | Bilateral regional free trade agreement | 2010/03 |
| China-Costa Rica | Bilateral regional free trade agreement | 2011/08 |
| China-Georgia | Bilateral regional free trade agreement | 2018/01 |
| China-Mauritius | Bilateral regional free trade agreement | 2021/01 |
- Data source: WTO Regional Trade Agreement Database.
- a Although China has signed FTAs with developed countries such as New Zealand and Australia, China's infrastructure investment projects in some developed countries are not reported in the data set, it is impossible to analyze the impact of FTA on the success or failure of such infrastructure projects in these countries. Therefore, there is no analysis of the impact of FTA on investment projects in developed countries.
- b In the China-ASEAN FTA, ASEAN include 10 countries: Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, Philippines, Singapore, Thailand, Viet Nam.
represents a country-level variable, including economic and political-level variables. For the economic level, we include gross domestic product(gdp), trade openness(tradeopen), and inflation(inflation) of the host country as control variables (Berger et al., 2013; Ismail et al., 2009; Yeyati et al., 2003). GDP reflects the economic development level of the host country. Higher GDP represents more market demand and can attract more foreign investment. Host country openness to trade is measured by the host country's total trade percentage of that year's GDP. Inflation of the host country is chosen to be the proxy for macroeconomic distortions. The political-level control variables include the quality of Institution(ql) and the host country's political stability(ps). Political factors influence investment behavior and affect the success rate of investment projects (Büthe & Milner, 2008; Desbordes & Vicard, 2009; Li et al., 2020). The host country rarely resorts to confiscation or nationalization. However, the hidden institutional risks still cause various losses to the home country enterprise's assets, and higher institution quality can better promote the smooth implementation of investment projects. Political stability is relevant to the willingness of foreign firms to invest in the host country. Besides, we also consider the project-level control variable. To check if the proposed Belt and Road Initiative influences the project, we introduce the variable BRI. If the start year of the project is after the proposal of BRI, the variable BRI is set to 1; otherwise, it is 0. We also included project scale (projectscale) measured by project monetary amount as the product control variable (Loncan, 2021). Project age influences the probability of project failure. The probability that a project will fail in the first 3 years may be very different from the probability of failure in 5 or 8 years. Recognizing the importance of project age, we add it as a control variable (projectage) and it is measured as the number of years from the project's start to year t. Type of financing is a product-level control variable (loan) that captures the effect of each type of financing (Tables 2 and 3).
| Variables | Definition | Data source |
|---|---|---|
| Represent whether the Chinese infrastructure investment project p failed in year t | “Global Chinese Development Finance Dataset” | |
| Dummy variable, set equal to1 in the case of a free trade agreement in force between China and host country, otherwise 0 | WTO Regional Trade Agreement Database http://rtais.wto.org/UI/PublicMaintainRTAHome.aspx | |
| gdp | Real GDP of the host country | World bank database, World Development Indicators |
| tradeopen | Sum of imports and exports of the host country in percent of GDP | World bank database, World Development Indicators |
| inflation | Inflation rate of the host country in percent (GDP deflator) | World bank database, World Development Indicators |
| ql | Reflects perceptions of the extent to which agents have confidence in and abide by the rules of society, and in particular, the quality of contract enforcement, property rights, the police, and the courts | World Governance Indicator |
| ps | measures perceptions of the likelihood of political instability and/or politically motivated violence | World Governance Indicator |
| BRI | Dummy variable, set equal to 1 if the implement year of the project is after 2013, else0 | “Global Chinese Development Finance Dataset” |
| projectscale | project monetary amount | “Global Chinese Development Finance Dataset” |
| projectage | Measured as the number of years from the start of the project to year t. | “Global Chinese Development Finance Dataset” |
| loan | Dummy variable for the type of project financing: 1 if a loan, and 0 otherwise | “Global Chinese Development Finance Dataset” |
| Observations | Mean | Std. dev. | Minimum | Maximum | |
|---|---|---|---|---|---|
| 4897 | 0.0240 | 0.1118 | 0.0000 | 1.0000 | |
| 4897 | 0.2078 | 0.4058 | 0.0000 | 1.0000 | |
| gdp | 4897 | 24.4833 | 1.9085 | 18.7137 | 28.6226 |
| tradeopen | 4897 | 0.8152 | 0.8861 | 0.1812 | 10.4301 |
| inflation | 4897 | 3.1920 | 0.9682 | −2.8026 | 7.2152 |
| ql | 4897 | −0.6592 | 0.5160 | −2.3212 | 1.3411 |
| ps | 4897 | −0.4884 | 0.7852 | −2.8123 | 1.3812 |
| BRI | 4897 | 0.6418 | 0.4795 | 0.0000 | 1.0000 |
| projectscale | 4897 | 18.3410 | 2.0679 | 7.7948 | 25.0808 |
| projectage | 4897 | 4.0957 | 3.0181 | 1.0000 | 20.0000 |
| loan | 4897 | 0.8531 | 0.3540 | 0.0000 | 1.0000 |
3 EMPIRICAL RESULTS
3.1 Baseline results
The empirical benchmark results are presented in Table 4, where the economy, politics of the host country, and project factors have been considered. We also considered the country pair, year, and industry fixed effects. The FTA, host country, and project control variables were gradually introduced in columns (1) to (3).
| Probit model | |||
|---|---|---|---|
| (1) | (2) | (3) | |
| FTA | −0.760** | −0.978*** | −1.253*** |
| (0.302) | (0.340) | (0.433) | |
| gdp | −0.145** | −0.187*** | |
| (0.057) | (0.069) | ||
| tradeopen | −0.087*** | −0.104*** | |
| (0.033) | (0.038) | ||
| inflation | 0.087 | 0.140** | |
| (0.066) | (0.069) | ||
| ql | −0.592*** | −0.890*** | |
| (0.208) | (0.269) | ||
| ps | −0.004 | −0.138 | |
| (0.122) | (0.135) | ||
| BRI | −1.302** | ||
| (0.569) | |||
| projectscale | −0.059 | ||
| (0.044) | |||
| projectage | 0.021** | ||
| (0.008) | |||
| loan | 0.518* | ||
| (0.298) | |||
| Country pair FE | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes |
| Industry FE | Yes | Yes | Yes |
| Observations | 5060 | 5014 | 4361 |
| Log likelihood | −332.50 | −309.35 | −237.52 |
- Note: *,**,*** Denote significance at the 1%, 5%, and 10% levels, respectively. The values in parenthesis underneath the estimated coefficients are standard errors.
The regression results showed that FTA reduces the failure rate of infrastructure projects in China's Belt and Road countries. Estimation results in column (3) suggested that FTA can reduce the failure rate of an infrastructure project. Economic-level control variables showed that the larger GDP and the higher the degree of openness to trade of the host country, the more conducive to the project's success. However, an increase in inflation is not beneficial for the project's success. The quality of the institution of the host country helps reduce the failure rate of the project, and the conclusion is consistent with our expectations. The host country's political stability can reduce the project's failure rate, although it is not significant at the 10% level. The proposal of the Belt and Road Initiative is conducive to improving the project's success rate. The reason is that the Belt and Road Initiative advocates interconnection, which can increase the investment confidence of both countries and promote the smooth progress of the project. The project scale has no significant impact on the success rate. A larger project can easily command the attention of the host country. However, it may lead to a higher probability that the underlying commercial contracts financed with Chinese loans and export credits will be renegotiated, which can cause an increase in uncertainty. We find that as project age increases, the rate of failure increases. During the implementation process, investors bear increasingly greater financial risks over time, leading to a higher probability of failure (Ikediashi et al., 2014; Kangari, 1988). Project financing through loans is associated with a higher rate of project failure compared to that through grants at the 10% significance level. Loans usually have higher investment costs than grants, thus increasing the risk of project failure (Kaming et al., 1997).
3.2 Robustness test
We further adopted a panel logit model to conduct the robustness test. The results are shown in columns (1)–(3) in Table 5. We found that FTA can significantly reduce the failure rate of an investment. At the same time, variables at other levels, including the host country's economy, politics, BRI, and project scale, have the same influence direction on the project's failure. These results are consistent with the results of the basic model, indicating that the results are highly robust.
| Logit model | IV estimation | |||||
|---|---|---|---|---|---|---|
| (1) | （2） | (3) | (4) | (5) | (6) | |
| FTA | −1.853*** | −2.336*** | −2.796*** | −0.662** | −0.954*** | −1.212*** |
| (0.650) | (0.707) | (0.779) | (0.285) | (0.306) | (0.368) | |
| gdp | −0.276** | −0.316*** | −0.101*** | −0.157*** | ||
| (0.108) | (0.100) | (0.030) | (0.046) | |||
| tradeopen | −0.177*** | −0.187*** | −0.0672*** | −0.087*** | ||
| (0.059) | (0.054) | (0.018) | (0.025) | |||
| inflation | 0.187 | 0.280** | 0.0633 | 0.121** | ||
| (0.132) | (0.129) | (0.0547) | (0.057) | |||
| ql | −1.161*** | −1.711*** | −0.430*** | −0.766*** | ||
| (0.383) | (0.369) | (0.119) | (0.162) | |||
| ps | −0.0251 | −0.327 | −0.006 | −0.137 | ||
| (0.243) | (0.252) | (0.085) | (0.110) | |||
| BRI | −2.386** | −1.075** | ||||
| (1.207) | (0.506) | |||||
| projectscale | −0.066 | −0.054 | ||||
| (0.078) | (0.034) | |||||
| projectage | 0.029** | 0.011** | ||||
| (0.013) | (0.005) | |||||
| loan | 1.143* | 0.468* | ||||
| (0.723) | (0.332) | |||||
| Country pair | Yes | Yes | Yes | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes | Yes | Yes | Yes |
| Industry FE | Yes | Yes | Yes | Yes | Yes | Yes |
| Observations | 5060 | 5014 | 4361 | 5060 | 5014 | 4361 |
| Log likelihood | −330.86 | −308.21 | −236.58 | −1773.68 | −1666.39 | −1465.89 |
| Wald test | 5.15** | 10.00*** | 10.71*** | |||
| AR test | 5.16** | 9.98*** | 10.67*** | |||
- Note: *, **, *** Denote significance at the 1%, 5%, and 10% levels, respectively. The values in parenthesis underneath the estimated coefficients are standard errors.
It can be argued that the estimated results presented so far are subject to endogeneity problems. There may be a certain reverse causality between FTA and investment success rate. That is, the more successful projects of foreign investment, the more the cooperation between the two parties may be further deepened, thus prompting the two parties to sign a trade agreement (Malesky, 2008). In addition, some other omitting variables may also lead to endogeneity. To reduce the effect of the potential endogeneity, we need to find an instrument variable.
An effective instrument must have two qualities: it must be a good predictor of the FTA but must not be correlated with the error term and hence the dependent variable. According to Mansfield (1998), the number of FTAs signed by countries other than country j but in the same geographic region can be a good instrument variable. This is because it can influence the probability of country j signing FTA but have no direct correlation with the foreign investment of country j. Hence, we identified the region for each country according to WTO region separation and counted the number of FTA in force for other countries that belonged to the same region with country j in each year.2 We found this measure to be a good predictor of the country's FTA (r = 0.24). The estimated results are reported in columns (4)–(6) of Table 5. The results confirm that FTA reduces the failure rate of infrastructure investments and promotes the successful implementation of infrastructure projects. The estimated results are qualitatively similar to those presented in Table 4, which implies that our main result is robust and accurate.
3.3 The depth of FTA
Considering that RTAs differ significantly in design, we included an “FTA depth” to reflect the FTA heterogeneity.
The depth data was obtained from the World Bank database. It measures depth in the case of legal constraints. First, it describes the depth of each provision and assigns its value in the following method: if the provision is not mentioned or is not legally binding, it is assigned 0; if the issue is clearly mentioned in FTAs and is legally binding but is excluded from the dispute settlement mechanism, it is assigned a value of 1; if the issue is clearly mentioned in the agreement, is legally mandatory, and applies to the dispute settlement mechanism, it is assigned a value of 2. Then, the depths of all provisions are summed and standardized to obtain the depth of FTAs. The value of FTA depth is between [0, 1]. Larger values represent deeper FTAs.
Columns (1)–(3) in Table 6 show the results, and we found that the deeper FTA agreement can better reduce the failure rate of infrastructure project investments.
| (1) | (2) | (3) | (4) | (5) | (6) | |
|---|---|---|---|---|---|---|
| FTAdepth | −1.511** | −2.275** | −2.513** | −1.441** | −1.882*** | −2.183*** |
| (0.755) | (1.015) | (1.068) | (0.568) | (0.650) | (0.757) | |
| gdp | −0.145** | −0.183** | −0.147** | −0.185*** | ||
| (0.059) | (0.073) | (0.057) | (0.069) | |||
| tradeopen | −0.089** | −0.106*** | −0.087*** | −0.103*** | ||
| (0.034) | (0.040) | (0.033) | (0.038) | |||
| inflation | 0.0893 | 0.143* | 0.087 | 0.139** | ||
| (0.067) | (0.073) | (0.066) | (0.069) | |||
| ql | −0.595*** | −0.867*** | −0.606*** | −0.886*** | ||
| (0.215) | (0.280) | (0.211) | (0.270) | |||
| ps | −0.021 | −0.096 | −0.001 | −0.137 | ||
| (0.127) | (0.144) | (0.123) | (0.134) | |||
| BRI | −1.380* | −1.299* | ||||
| (0.719) | (0.670) | |||||
| projectscale | −0.070 | −0.060 | ||||
| (0.048) | (0.044) | |||||
| projectage | 0.022** | 0.019** | ||||
| (0.008) | (0.009) | |||||
| loan | 0.512 | 0.514* | ||||
| (0.323) | (0.297) | |||||
| Country pair FE | Yes | Yes | Yes | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes | Yes | Yes | Yes |
| Industry FE | Yes | Yes | Yes | Yes | Yes | Yes |
| Observations | 5060 | 5014 | 4361 | 5060 | 5014 | 4361 |
| Log likelihood | −335.49 | −313.18 | −243.61 | −332.30 | −308.97 | −238.10 |
- Note: *,**,*** Denote significance at the 1%, 5%, and 10% levels, respectively. The values in parenthesis underneath the estimated coefficients are standard errors.
Then we performed a robustness test by changing the FTA depth calculation method to verify the conclusion's stability. We followed Dür et al. (2014) to obtain the FTA depth. Considering that not all agreements in FTA are equally important to national commitments, latent trait analysis was used to process highly relevant data to obtain more accurate results. A total of 49 theoretically related variable agreements were used (These variables involved some major areas, such as service liberalization, trade-related investment measures, intellectual property rights, standards, public procurement, and competition policy) to measure depth. We also standardized the depth value and made it between [0, 1]. The results are shown in columns (4)–(6) in Table 6. We concluded that FTA depth still has a significant negative effect on the failure rate of project investment, and the coefficient is relatively close to the results in columns (1)–(3). This further proves that the positive effect of deepening FTA on the success rate of infrastructure investment is robust.
4 EXPLORE KEY PROVISIONS INSIDE THE BLACK BOX OF FTA
The “Global Chinese Development Finance Dataset” tells the story of each project and provides opportunities for us to investigate why BRI infrastructure projects failed. Through statistical analyses, we found that among the failed projects, 77% recorded underperformances and inefficiency problems. For example, infrastructure projects encountered resistance during execution, and contractors failed to meet key milestones specified in their contracts3. As investment inefficiency has led many projects to fail, it is important to check whether investment provisions in the FTA will reduce the probability of failure. We refer to Berger et al. (2013) and considered national treatment in the pre-establishment phase and investment dispute settlement mechanism clauses. They refer to investment liberalization and protection and constitute an important legal guarantee.
Table 7 shows the results. First, we considered if the FTA included investment provisions; we used FTAwithinvest and FTAwithoutinvest to distinguish. The results in column (1) show that FTA containing the investment protocols has an important effect on reducing failure fate while FTA without investment protocols has no significant effect.
| (1) | (2) | (3) | |
|---|---|---|---|
| FTAwithinvest | −1.453*** | ||
| (0.455) | |||
| FTAwithoutinvest | 0.211 | ||
| (0.710) | |||
| FTAwithNT | −1.605*** | ||
| (0.532) | |||
| FTAwithoutNT | −0.230 | ||
| (0.491) | |||
| FTAwithISDS | −1.674*** | ||
| (0.532) | |||
| FTAwithoutISDS | 0.617 | ||
| (0.460) | |||
| gdp | −0.185*** | −0.191*** | −0.200*** |
| (0.062) | (0.065) | (0.064) | |
| tradeopen | −0.102*** | −0.102*** | −0.100*** |
| (0.034) | (0.035) | (0.034) | |
| inflation | 0.142** | 0.147** | 0.161** |
| (0.066) | (0.068) | (0.068) | |
| ql | −0.843*** | −0.888*** | −0.918*** |
| (0.229) | (0.245) | (0.237) | |
| ps | −0.148 | −0.153 | −0.206 |
| (0.127) | (0.127) | (0.133) | |
| BRI | −1.262** | −1.303** | −1.347** |
| (0.616) | (0.638) | (0.634) | |
| projectscale | −0.057 | −0.060 | −0.069 |
| (0.041) | (0.043) | (0.043) | |
| projectage | 0.018** | 0.019** | 0.019** |
| (0.008) | (0.009) | (0.009) | |
| loan | 0.540* | 0.508* | 0.540* |
| (0.283) | (0.286) | (0.289) | |
| Country pair FE | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes |
| Industry FE | Yes | Yes | Yes |
| Observations | 4361 | 4361 | 4361 |
| Log likelihood | −235.59 | −235.45 | −234.16 |
- Note: *, **, *** Denote significance at the 1%, 5%, and 10% levels, respectively. The values in parenthesis underneath the estimated coefficients are standard errors.
In column (2), we further considered whether FTA included investment national treatment in the pre-establishment phase. FTAwithNT means that China and the host country j have an FTA agreement in force in year t and include the investment national treatment in the pre-establishment phase. FTAwithoutNT means that China and the host country j have signed an FTA agreement without the investment national treatment protocol. We found that FTA with NT positively impacted the project success rate, at the 1% level of significance, while FTA without NT has no significant impact. The use of pre-establishment NT in the FTA investment agreement can lower the threshold for foreign investment, reduce approval items and optimize the handling process. This also means that the host country creates a policy environment that treats foreign investment equally.
In column (3), we distinguished whether FTA includes an investment dispute settlement mechanism. FTAwithISDS indicates that China and the host country j have an FTA agreement in force in year t and include the investment dispute settlement mechanism in the FTA provisions. In contrast, FTAwithoutISDS does not include a dispute settlement mechanism. It shows that FTA that contains investment dispute settlement mechanism also plays a positive role in the project's successful implementation, while FTAwithoutISDS has no significant effect. Using an investment dispute settlement mechanism can provide compulsory protection for foreign direct investment. It protects every aspect of investments in terms of treatment after enterprise entry, repatriation of asset returns, exemption from expropriation and nationalization, and compensation for losses. It can effectively protect various assets and income invested in the host country, thereby greatly reducing the risk and cost of investment.
These results are meaningful. From the host country's perspective, the provisions represent the government's commitment to investment. It also promotes investment liberalization and provides investment protection in the form of legal provisions.
From the Chinese government's perspective, the higher standard of bilateral investment treaty can better protect the interests of the Chinese overseas investment projects. However, the high standard also presents higher requirements for investment management system and market openness. High-standard investment agreements require further refinement of relevant provisions in investment agreements, such as increasing operability in terms of the scope of investment definitions and specific standards of national treatment. These are all conducive to improving investment efficiency and enhancing investment success rate.
5 CONCLUSION
This paper analyzes whether the signing of FTA agreements is conducive to promoting the successful implementation of China's infrastructure investment projects in the countries along the Belt and Road. It is concluded that the signing of the FTA agreement is conducive to reducing the failure rate of China's investment in the countries along the Belt and Road. For deeper FTAs, the effect is more pronounced. Through further analysis of failed investment projects, we find that most projects fail due to inefficient execution. Therefore, this paper further explores the investment provisions of FTA and concludes that investment provision can improve the success rate of investment. Specifically, the investment national treatment in the pre-establishment and the investment dispute settlement mechanism can greatly improve the investment efficiency and guarantee the smooth implementation of the project.
The conclusions of this paper have important policy implications for the study of OFDI in China.
On the one hand, China should enter into FTA agreements with countries along the Belt and Road, and establish more trade partnerships to stimulate foreign investment.
On the other hand, in negotiating the FTA agreements, the Chinese government must continue to update and improve the provisions of bilateral agreements in accordance with economic and social development processes. The government especially needs to pay attention to the liberalization and guarantee mechanism in the investment agreement and promote the agreement to a high standard.
As a strategy in outward foreign investment, investors should also prioritize host countries that have signed free trade agreements with China. In particular, these are countries that have signed investment protection clauses with China, such as the Treatment of Investments (Pre-Establishment) and Investment Dispute Settlement Mechanisms.
There are also limitations in the availability of information in the data set. For example, it does not include private investment projects. Hence, this study only examines the impact of FTA on the rate of success or failure of investment projects financed by governments and public institutions. In future, studies should systematically investigate the impact of FTA on all types of projects.
AUTHOR CONTRIBUTIONS
Beibei Hu: Conceptualization; Data curation; Methodology; Project administration; Writing original draft; Writing review editing. Yuying Jin: Project administration; Supervision; Writing—review & editing. Kai Wang: Conceptualization; Formal analysis; Investigation; Resources; Writing—review & editing.
ACKNOWLEDGMENTS
The authors acknowledge financial support from the Fundamental Research Funds for the Central Universities (2020110308), the Ministry of Education of Humanities and Social Science Project (20YJC790043), and Project of Shanghai “Science and Technology Innovation Action Plan 2022” (22692105200).
CONFLICTS OF INTEREST
The authors declare no conflicts of interest.
ETHICS STATEMENT
Not applicable.
REFERENCES
- 1 The Belt and Road Initiative (BRI) is China's long-term policy and investment program that integrates the Silk Road Economic Belt (Belt) and 21st-Century Maritime Silk Road (Road) into one initiative. It was proposed in 2013 with 65 Asian, European, and African countries participating. As more countries became interested in the BRI, its definition has broadened. By the end of 2021, this number grew to 149, constituting mostly developing countries. (Source: https://www.yidaiyilu.gov.cn/).
- 2 The region can be divided into Asia, Europe, Africa, North America, South America, and Ocean.
- 3 We identify these projects by applying the following keywords: inefficiency, underperform-ance, freeze, losses, failure, threat, halt, suspend,bankrupt, reschedule,shutter,delay, overrun, default, sol-vency, arrears, rescind, remedia, jeopardy, abandon, renegotiate, mothball, withheld, halt,seize. We then reviewed each of the project description by hand to remove false positives.
