Global value chain integration and productivity: the case of Turkish manufacturing firms

ABSTRACT

In this study, we examine how firms’ positions (supplier, final, or both) in both global and domestic value chains (GVC and DVC) affect their productivity. This is said to be the first attempt in exploring the impact of the integration of firms on the GVCs on productivity generation in the Turkish manufacturing industry at the firm-level. The analysis is based on firm-level data obtained from the Turkish Statistical Institute (TurkStat) and covers the period from 2003 to 2015. The data used in the analysis includes all firms employing 20 or more employees in the Turkish manufacturing industry. Our findings based on both fixed-effects and GMM estimations show that while supplier position on the domestic chain has a negative effect on productivity, the same position in GVC vanishes this effect. The final firm position in the GVC, on the other hand, provide more benefits to SMEs than to large-scale firms.

KEYWORDS:

• GVC
• DVC
• productivity
• Turkish manufacturing firms

Subject classification codes:

• D24
• E24
• J24
• L60
• O14

Acknowledgments

We thank the Turkish Statistical Institute (TurkStat) for providing the firm-level data on production and trade, namely Annual Industry and Service Statistics and Trade Statistics. All analyses have been conducted at the Microdata Research Centre of TurkStat in Ankara, Turkey with respect to the protocol on the statistic secret and personal data protection. The results and the interpretations expressed in this study are the exclusive responsibility of the authors, and by no means represent official statistics of TurkStat.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 TurkStat does not permit the database to be removed from its premises due to data confidentiality. Thus, all empirical analyses in this project were conducted at the Micro Data Research Center of TurkStat in Ankara, Turkey based on a Protocol of data confidentiality and data security. TurkStat allows researchers to take the results of their analysis out after controlled by related Departments of TurkStat.

2 See Appendix A for the classification of industries according to their technology intensity.

3 We used a perpetual inventory method to calculate capital stock of by using gross fixed capital formation, depreciation expenses information obtained from the surveys. Following Taymaz et al. (2008) and Kılıçaslan et al. (2017), firstly capital stock for initial of each firm year must be calculated by dividing depreciation value to depreciation rate. Then current investment is added with previous year capital stock adjusted with depreciation.

4 To identify whether the firm purchases intermediate goods, we used ‘semi-finished goods’ and ‘parts & components’ stages of BEC classification developed by Gaulier et al. (2007).

5 This interaction term has been found insignificant in most of specifications. Therefore, we did not interpret this variable henceforth.

6 Also, because of very low observations in high-tech DVC firms, we could not estimate any coefficient.

Additional information

Funding

This work was supported by the Economic Research Forum (ERF), Giza, Egypt under the project ‘Structural Change, Resource Misallocation and Growth Dynamics in the MENA Region.’ The views expressed in this paper are those of the authors and do not necessarily represent those of the ERF.