Skip to main content
Log in

Containing Chinese state-owned enterprises? The role of deep trade agreements

  • Original Paper
  • Published:
Review of World Economics Aims and scope Submit manuscript

Abstract

Regional trade agreements increasingly include provisions that regulate state-owned enterprises. This paper combines new information on the content of “deep” regional trade agreements and data on Chinese firm-level exports during 2000–2011 to analyze the spillover effect of rules on state-owned enterprises on the intensive and extensive margins of Chinese state-owned enterprises’ trade. Rather than containing state capitalism, regional trade agreements regulating state-owned enterprises signed by Chinese trading partners with third countries increase exports and entry of Chinese state-owned enterprises as they gain a competitive edge in regulated markets. This spillover effect is robust to several extensions and is even stronger for agreements that include rules on subsidies and competition policy. This finding points to the need for commonly agreed multilateral rules to regulate state owned enterprises.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others

Notes

  1. As we will discuss in the next section, many of the trade agreements regulating SOEs were signed post 2000 and are between developed and developing countries. Examples include EU-South Africa (2000), Japan-Indonesia (2008), USA-Panama (2012).

  2. RTAs typically reduce trade costs between members, but their impact on trade costs with non-members is determined by the precise nature of the provisions contained in the agreements. Mattoo et al. (2022) build on an intuition by Baldwin (2014) and point out that deep trade agreements lead to more trade creation and less trade diversion than shallow agreements. Some provisions are purely discriminatory, while some others (e.g. subsidies, competition) do not discriminate between members and non-members of an RTA. Thus, some provisions reduce trade costs between members and increase trade costs to non-members, while others can reduce trade costs for both members and non-members.

  3. Enforceable rules include those rules for which the language of the agreement is sufficiently precise from a formal legal point of view and where the agreement foresees a dispute settlement mechanism to resolve disagreement.

  4. An exception is the recent work by Baccini et al. (2019) that investigates the role of the entry of Vietnam in the World Trade Organization on the performance (i.e. selection and productivity) of Vietnamese SOEs relative to their private counterparts.

  5. Trade diversion is however not systematic. The imports of non-members have not been affected by the integration of Turkey in the European Community (Magee, 2015) or three RTAs signed in Latin America (Freund, 2010)

  6. The comprehensive list of chapters includes Anti-dumping, Competition policy, Countervailing duties, Environment, Export taxes, Investment, Intellectual property rights, Labor market, Migration, Movement of capital, Public procurement, Rules of origins, Services, Sanitary and Phytosanitary, State-owned enterprises, Subsidies, Technical barriers to trade and Trade facilitation.

  7. The data can be accessed at http://datatopics.worldbank.org/dta/index.html.

  8. Exceptions include trade agreements with ASEAN (2005), Singapore (2009), Iceland (2014), Switzerland (2014) and the republic of Korea (2015).

  9. Appendix A” provides the details regarding the construction of the depth variables.

  10. Key SOE provisions are listed in “Appendix B”.

  11. This database presents the advantage of reconciling importer and exporter declarations through a harmonization procedure.

  12. Data can be accessed at https://fortune.com/global500/2019/.

  13. We aggregate BACI data at the exporter-importer-time level and extract the exported value between 1995 and 2015.

  14. Importer–exporter fixed effects attenuate the concerns regarding the endogeneity of trade agreements by controlling for countries’ predispositions to form an agreement (Baier & Bergstrand, 2007).

  15. We measure the overall depth of a given RTA based on the number of provisions included in the agreement across all the policy areas. While this is a rough measure of overall depth, the underlying idea is that a larger number of provisions reflects the extent of the liberalizing and enforcement obligations that members are committing to.

  16. In order to have an idea of the absolute effects (rather that relative), “Appendix C” estimates the impact of signing SOE provisions with third countries on Chinese exports by controlling for demand through destination specific GDP instead of destination-time specific fixed effect. The results show a positive and significant coefficient for the multilateral depth of SOE provisions.

  17. Appendix D” estimates the coefficient of multilateral depth of enforceable SOE provisions across countries and shows that China appears as having a large coefficient compared to other countries. The Appendix reports the results for countries having a large population (i.e. above 25 million) for clarity.

  18. Another potential explanation is that domestic SOEs are less competitive and begin to import cheaper inputs coming from China. If this is the only explanation, we should see a positive impact only for intermediate products. “Appendix E” and “Appendix F” show that this impact is robust for Chinese exports of intermediates and non-intermediates products.

  19. Precise computation of the trade impact of RTAs is detailed in “Appendix G

  20. Precise computation of the spillover impact of RTAs is detailed in “Appendix H”.

  21. A concern about the high correlation between the depth of SOE provisions and the depth of the agreement is alleviated by our fixed effect structure. Accounting for those fixed effects reduces the correlation between the two variables to 0.14.

  22. \( \frac{35}{1459} \times 0.998 = 0.024\), with 1,459 the maximum number of multilateral SOE provisions and 0.998 the estimated coefficient.

  23. \(\frac{35}{1459} \times 1.866 = 0.045\), with 1,459 the maximum number of multilateral SOE provisions and 1.866 the estimated coefficient.

  24. The concern that unobserved components of the export transactions for each firm across all its destination markets could be correlated remains. We address this concern by estimating our micro-level benchmark specification and report standard errors that account for clustering at the firm level. Results are reported in “Appendix I” and do not affect our benchmark results.

  25. Note that spillover effects may also work through Global Values Chains (GVCs). If two country members of an RTA are also part of a GVC network involving Chinese firms, the estimated spillover effects could in part reflect sector-specific GVC links with China. Because our strategy compares the impact of Chinese SOEs with respect to Chinese private firms, GVC links would have to be stronger for Chinese SOEs with respect to Chinese private firms to explain our results. The data for two key industries in the sector-specific results, machinery/electronics and textile shows that Chinese SOEs are responsible for less than 9% and 25% of GVC exports, respectively. This indicates that GVC links are unlikely to drive our results.

  26. To motivate the focus on subsidies and competition policy, we run the regressions (columns 1 and 5, Table 3) for all policy areas covered by the deep trade agreements database. Results show that -as expected- the spillover effect associated to SOE rules has larger point estimates. Provisions related to subsidies and competition policy also stand out, consistently with intuition. Other policy areas appear to provide relevant spillovers. Namely, deeper commitments on rules of origin and countervailing duties in third-country agreements boost exports of Chinese SOEs relative to private firms in members’ markets, while the opposite is true for trade facilitation provisions. The figures in “Appendix J” have the details.

  27. We cannot investigate the direct impact of subsidy and competition provisions at the firm-level as China does not have any agreement including those provisions. Examples of subsidy and competition provisions extracted from our database are respectively listed in “Appendix B”.

  28. We focus on the impact at the intensive margin since this procedure requires a lot of computational power.

  29. \(\frac{1.331 \times 0.799}{(1.331 \times 0.799 + 0.751 \times 0.201)}\times 100= 87.5\), with 1.331 the estimated coefficient of multilateral depth of enforceable SOE provisions for new exporters, 0.751 the estimated coefficient of multilateral depth of enforceable SOE provisions for incumbents, 0.799 the share of total exports coming from new exporters and 0.201 the share of total exports coming from incumbents.

  30. Note that all SOE provisions are binary.

References

Download references

Acknowledgements

We are grateful to Matthieu Crozet, Ana Cusolito, Ohyun Kwon, Mario Larch, Petros Mavroidis, Martin Raiser, Luca Rubini, Costas Syropoulos, Yoto Yotov and seminar participants at the World Bank webinar series on "The Economics of Deep Trade Agreements", Columbia University Law School, Drexel University, CEPII, University of Paris Saclay, ETH Zurich for helpful comments and suggestions. Research for this paper has been in part supported by the World Bank’s Multidonor Trust Fund for Trade and Development. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Kevin Lefebvre.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendices

Appendices

1.1 Appendix A: Construction of depth variables

We combine two different types of data to construct the indexes of depth. The first one is the Deep Trade Agreement Database [i.e. see Mattoo et al. (2020)], produced by the World Bank. It covers all the provisions included in 283 trade agreements, with detailed information on different chapters (Technical Barrier to Trade, Sanitary and Phytosanitary measures, State-Owned Enterprises, Subsidies, Environment, Services, Competition policy, etc.). While most of the individual provisions are coded as yes/no questions, some are coded as multiple choice. In order to construct consistent indexes of depth, we then restrict our indexes of depth to provisions that are binary.Footnote 30 In a second step, we attribute time-varying bilateral country pairs to every trade agreement. Note that to capture the evolving nature of the European Union, each provision signed between the EU and a third country is automatically included in intra-EU bilateral pairs. Finally, each bilateral country pair relationship is extended for all the periods in which it is in force. Note that a country pair having several trade agreements increases the number of provisions in force over time only if these agreements include different provisions. From this database, we can then construct the number of SOE provisions in force between exporting country i and importing country j at time t (i.e. Bilateral depth of SOE), the total number of provisions excluding the one related to SOEs in force between exporting country i and importing country j at time t (i.e. Bilateral depth of RTA), the number of SOE provisions in force between importing country j and all its trade partners except country i at time t (i.e. Multilateral depth of SOE) and the total number of provisions excluding the one related to SOEs in force between importing country j and all its trade partners except country i at time t (i.e. Multilateral depth of RTA).

The second one is a specific database constructed by Rubini and Wang (2020) coding each provision related to SOEs. The database provides detailed information on the enforceability and whether each SOE provision goes beyond the commitments signed at the WTO. SOE provisions are enforceable when the language is considered binding with state-to-state dispute settlement or private dispute settlement or both. We merge the two data sets and construct the bilateral and multilateral indexes of depth for enforceable SOE provisions, enforceable & WTO+ SOE provisions and enforceable key SOE provisions.

1.2 Appendix B: Key SOE, subsidy and competition provisions in RTAs

SOE provisions

 Does the agreement regulate ownership or property regimes, or liberalization processes?

 Does the agreement prohibit discrimination by state enterprises?

 Does the agreement regulate subsidization to state enterprises?

 Does the agreement prohibit anti-competitive behaviour of state enterprises?

Subsidy provisions

 Does the agreement prohibit or regulate export subsidies?

 Does the agreement prohibit or regulate subsidies distorting trade or competition?

 Does the agreement include any national treatment obligation (goods) for subsidies?

 Does the agreement provide for a dispute settlement mechanism to deal with subsidy issues?

 Does the agreement provide for any institution to deal with transparency or enforcement?

 Does the agreement regulate the imposition of countervailing duties?

Competition provisions

 Does the agreement prohibit/regulate cartels/concerted practices?

 Does the agreement prohibit/regulate abuse of market dominance?

 Does the agreement regulate undertakings with exclusive rights?

 Does the agreement regulate monopolies?

 Does the agreement regulate anti-competitive behaviour of SOEs?

 Does the agreement regulate state aid?

 Does the agreement regulate mergers and acquisitions?

1.3 Appendix C: Country-level impact of SOE provisions

 

\(X_{ijt}\)

\(X_{ijt}\)

\(X_{ijt}\)

\(X_{ijt}\)

 

All

Enf.

\(Wto^+_{enf}\)

\(Key_{enf}\)

\(\ln (GDP_{jt})\)

1.053***

1.083***

1.082***

1.092***

 

(0.0430)

(0.0426)

(0.0427)

(0.0426)

Bilateral RTA (dummy)

− 0.431***

− 0.421***

− 0.442***

− 0.489***

 

(0.0925)

(0.0876)

(0.0790)

(0.0893)

Bilateral depth of RTA

− 0.311

− 0.288

− 0.183

0.782

 

(0.727)

(0.657)

(0.606)

(0.689)

Bilateral depth of SOE

0.761***

1.246***

1.391***

0.398

 

(0.229)

(0.315)

(0.344)

(0.237)

Multilateral RTA (sum)

− 1.022**

− 0.448

− 0.336

− 0.648*

 

(0.355)

(0.322)

(0.327)

(0.321)

Multilateral depth of RTA

0.0397

0.122

0.0244

0.124

 

(0.250)

(0.325)

(0.332)

(0.289)

Multilateral depth of SOE

1.147***

0.511***

0.540***

0.925***

 

(0.287)

(0.134)

(0.135)

(0.144)

Fixed-effects

it ij

it ij

it ij

it ij

SE

Robust

Robust

Robust

Robust

N

3897

3897

3897

3897

  1. *, ** and *** indicate significance at 5%, 1%, and 0.1% levels, respectively

1.4 Appendix D: SOE regulation spillovers across countries

figure a

1.5 Appendix E: Country-level impact of SOE provisions on export of intermediate products

 

Export value of intermediate products

 

All

Enf.

\(Wto^+_{enf}\)

\(Key_{enf}\)

Bilateral RTA (dummy)

− 0.0306

− 0.0364*

− 0.0353*

− 0.0279

 

(0.0157)

(0.0159)

(0.0161)

(0.0162)

Bilateral depth of RTA

0.331***

0.246***

0.240***

0.298***

 

(0.0405)

(0.0479)

(0.0478)

(0.0358)

Bilateral depth of SOE

− 0.0752

0.0307

0.0391

− 0.0529*

 

(0.0410)

(0.0466)

(0.0474)

(0.0254)

CHN \(\times\) Multilateral RTA (sum)

− 0.0443

0.240

0.252

0.195

 

(0.202)

(0.201)

(0.205)

(0.188)

CHN \(\times\) Multilateral depth of RTA

− 0.343

− 0.373

− 0.381

− 0.466*

 

(0.186)

(0.207)

(0.213)

(0.182)

CHN \(\times\) Multilateral depth of SOE

0.459*

0.206*

0.213*

0.507***

 

(0.201)

(0.0984)

(0.106)

(0.0956)

Fixed-effects

it jt ij

it jt ij

it jt ij

it jt ij

SE

robust

robust

robust

robust

N

598,304

598,304

598,304

598,304

  1. *, ** and *** indicate significance at 5%, 1%, and 0.1% levels, respectively

1.6 Appendix F: Country-level impact of SOE provisions on export of non-intermediate products

 

Export value of non-intermediate products

 

All

Enf.

\(Wto^+_{enf}\)

\(Key_{enf}\)

Bilateral RTA (dummy)

− 0.173***

− 0.147***

− 0.150***

− 0.185***

 

(0.0303)

(0.0245)

(0.0246)

(0.0286)

Bilateral depth of RTA

0.114*

0.315***

0.337***

0.203***

 

(0.0489)

(0.0579)

(0.0582)

(0.0449)

Bilateral depth of SOE

0.231***

− 0.0504

− 0.0771

0.169***

 

(0.0654)

(0.0470)

(0.0483)

(0.0342)

CHN \(\times\) Multilateral RTA (sum)

− 0.901***

− 0.378

− 0.253

− 0.794***

 

(0.265)

(0.236)

(0.241)

(0.238)

CHN \(\times\) Multilateral depth of RTA

0.967***

0.308

0.0704

0.363

 

(0.220)

(0.264)

(0.274)

(0.211)

CHN \(\times\) Multilateral depth of SOE

0.544*

0.832***

1.014***

1.414***

 

(0.260)

(0.157)

(0.164)

(0.164)

Fixed-effects

it jt ij

it jt ij

it jt ij

it jt ij

SE

Robust

Robust

Robust

Robust

N

589,349

589,349

589,349

5893,49

  1. *, ** and *** indicate significance at 5%, 1%, and 0.1% levels, respectively

1.7 Appendix G: Quantification of the trade impact of RTAs

Direct impact: trade impact between members

  • EU-COL/PER agreement (2013)

  • 341 provisions

  • 12 enforceable SOE provisions

\(1 \times (-0.0387) + 341 \times \frac{1}{604} \times 0.234 + 12 \times \frac{1}{17} \times 0= 9.3\%\)

  • EU-KOR agreement (2011)

  • 345 provisions

  • 0 enforceable SOE provisions

\(1 \times (-0.0387) + 345 \times \frac{1}{604} \times 0.234 + 0 \times \frac{1}{17} \times 0= 9.5\%\)

1.8 Appendix H: Quantification of the spillover impact of RTAs

Indirect impact: Chinese exports to members

  • EU-COL/PER agreement (2013)

  • 341 provisions

  • 12 enforceable SOE provisions

\(1 \times \frac{1}{84} \times 0 + 341 \times \frac{1}{26783} \times 0 + 12 \times \frac{1}{654} \times 0.541= 1\%\)

  • EU-KOR agreement (2011)

  • 345 provisions

  • 0 enforceable SOE provisions

\(1 \times \frac{1}{84} \times 0 + 345 \times \frac{1}{26783} \times 0 + 0 \times \frac{1}{654} \times 0.541= 0\%\)

1.9 Appendix I: Firm-level impact of SOE provisions with clustered standard errors

 

Participation

\(\ln (X_{fjt})\)

 

Enf.

Enf.

soe \(\times\) Bilateral TA (dummy)

− 0.0667***

− 0.219***

 

(0.00250)

(0.0303)

soe \(\times\) Bilateral depth of TA

0.168***

0.331***

 

(0.00639)

(0.0761)

soe \(\times\) Bilateral depth of SOE

− 0.0746***

0.0287

 

(0.00417)

(0.0471)

soe \(\times\) Multilateral TA (sum)

− 0.0197*

− 0.233*

 

(0.00921)

(0.112)

soe \(\times\) Multilateral depth of TA

− 0.120***

− 0.0305

 

(0.0103)

(0.124)

soe \(\times\) Multilateral depth of SOE

0.110***

0.445***

 

(0.00792)

(0.0951)

Fixed-effects

fj ft jt

fj ft jt

Cluster

f

f

N

40,922,880

8,390,390

r2

0.626

0.796

  1. *, ** and *** indicate significance at 5%, 1%, and 0.1% levels, respectively

1.10 Appendix J: Firm-level impact of other chapters’ spillovers

See Figs. 10 and 11.

Fig. 10
figure 10

Firm-level impact of other chapters’ spillovers—Extensive margin

Fig. 11
figure 11

Firm-level impact of other chapters’ spillovers—Intensive margin

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Lefebvre, K., Rocha, N. & Ruta, M. Containing Chinese state-owned enterprises? The role of deep trade agreements. Rev World Econ 159, 887–920 (2023). https://doi.org/10.1007/s10290-022-00484-z

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10290-022-00484-z

Keywords

JEL Classification

Navigation