Incomplete contracts and allocation of residual claim along the global value chains: evidence from Japan

ABSTRACT

The seminal work by Antràs and Chor in 2013 predicted that, under incomplete-contracting environments, the optimal allocation of residual claim for firms participating in global production depends on the relative position at which the supplier enters the sequential production and on the demand elasticity faced by the final-good producer. Using data of Japanese value-added exports, this study empirically verifies this theoretical prediction through supportive evidence.

KEYWORDS:

• global value chains
• incomplete contracts
• residual claim
• Japan

JEL CLASSIFICATION:

• F14
• F23
• D23
• L20

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Empirically, AC correlated the share of the U.S. intra-firm imports in total U.S. imports reported during the period 2000–2010, that is, the measure of integration decisions, with the supplier’s position of that industry. In the complements (industries with above-median average demand elasticity), the average U.S. intra-firm import share decreases with the upstreamness measure. However, in the substitutes cases (industries with below-median average demand elasticity), the intra-firm trade share increases across terciles of its upstreamness.

² The overseas assets of Japan in 2017 were around 1,012 trillion Japanese Yen (JPY) (around twice its domestic gross production in 2017), with net assets of 328 trillion JPY (Data source: The Ministry of Finance of Japan, 2018).

³ The industry classification follows the international standard industrial classification of all economic activities (ISIC) Revision 3.1 classification for the non-EU countries, which is compatible with the Nomenclature generale des Activites economiques dans les Communautes europeennes (NACE) Revision 1.1 that is used for EU countries.

⁴ Taking the global production chain of iPhone as an example, Xing and Detert (2010) used dvas to measure the real gains of Chinese manufacturers. Based on their estimates, Chinese workers contribute only $6.5 (i.e., the dva) to each iPhone (about 3.6% of the total manufacturing cost). This demonstrates that the organizer of the iPhone global production chain has allocated a very small share of residual claim along the supplier in China.

⁵ Production length is defined as number of times that the value-added is counted from initial inputs in one sector to final goods in another sector. Splitting the total abroad production length into Richard’s trade and GVCs’ related trade, GVCs’ production line position (i.e., up_{country i,j,t}) can be obtained by dividing GVCs’ production length based on the forward linkage by GVCs’ production length based on the backward linkage. It is apparent that the larger the index, when compared to the backward angle, the more would be steps that will be needed for an exported original domestic converted into final products consumed by foreign customers inputs in forward angle, which apparently represents more upstream in GVCs.

⁶ As data of cross-input substitutability are not readily available, we follow AC by assuming that this indicator, among the stage inputs, is largely uncorrelated with the elasticity of demand, $\rho$, faced by the average buyer of an industry’s output. In this manner, we associate the sequential complements case with high values of $\rho$ and the substitutes’ case with low values of $\rho$.

⁷ The selection of control variables follows AC.

⁸ We also apply the latest WIOD revision (Release 2016) to ensure that our baseline results are not driven by the specific collection of data. This allows for estimating our econometric specifications with a larger sample size (41 countries/regions * 18 industries * 15 years); the main results come to be highly consistent with our baseline estimates. Results are available upon request.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [grant numbers: 71363016, 71663023, 71703034, 71703042, 71863010].