On tariff changes and firm-production evolution: insights from Turkish manufacturing

Abstract

We contribute to the yet limited evidence on the relationship between trade liberalisation and a firm's product mix and diversification strategies for an emerging economy, Turkey. Lower import barriers foster firms’ specialisation in their core products. A drop in import tariffs, indeed, enhances a firm's propensity to drop fringe varieties and favours production growth of core products. More importantly, it favours firms’ specialisation in more sophisticated goods. Export tariff cuts, instead, by relaxing competitive pressure at home and lowering the cost to export, only reduce the firms’ incentive to innovate.

Keywords:

• product mix
• production diversification
• import tariffs
• export tariffs

JEL Classifications:

• L25
• F14
• D21

Acknowledgements

The data used in this work are from the Foreign TradaData, the Annual Business Statistics and the Production Surveys provided by Turkish Statistical Office (Turkstat). All elaborations have been conducted at the Microdata Research Centre of Turkstat in Ankara under the respect of the law on the statistic secret and the personal data protection. The results and the opinions expressed in this article are exclusive responsibility of the authors and, by no means, represent official statistics. We are grateful to TurkStat staff from foreign trade statistics, SBS and dissemination departments for their help.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Bernard, Redding, and Schott (2010) for the USA and Goldberg et al. (2010b) for India show that although the intensive margin – changes of existing products’ output – is the main driver of output growth of continuing firms in manufacturing, their extensive margin – changes in firm product mix – is quantitatively significant. For Chilean firms the latter is, instead, the most important component of manufacturing output growth (Navarro 2012). Empirical evidence, thus, confirms that the within firm-product churning can significantly contribute to the economic development and, even in case of an unchanged basket of produced goods, the existing products’ expansion and contraction is not an unimportant matter. Also, multi-product firms account for a relevant share of a country's manufacturing sector and changes in their product mix constitute an essential part of industry dynamics, both in developed and developing economies (Navarro 2012; Bernard, Redding, and Schott 2010).

2. The dynamism of the Turkish economy has been proved by Hidalgo (2009) who showed that Turkey, together with Brazil and Indonesia, was one of the few countries that dramatically increased its product space's complexity over the period 1963–2005.

3. The first eight digits of PRODTR codes correspond to PRODCOM codes, the first six digits to CPA codes and the first four digits to NACE codes.

4. WITS-TRAINS data-set provide us with the effectively applied tariffs by countries. Then if a preferential tariff exists, it will be used as the effectively applied tariff.

5. From the comparison of Table 1 and the similar statistics in Bernard, Redding, and Schott (2010), we could erroneously conclude that production in Turkey is less concentrated than in the United States. The classification used by Bernard, Redding, and Schott (2010) is however more aggregated than the one in this paper, thus preventing any direct comparison.

6. Similar results on the baseline specification were obtained when the Herfindahl index was used, nonetheless to provide a methodology that is in line with extant work on firm production choices we decided to stick to the entropy index.

7. We define as exporter those firms selling own produced goods to foreign customers, thus excluding from our definition firms acting as trade intermediary in foreign markets for other Turkish producers.

8. We alternatively tested tariff impacts on the two main firm's products and we got substantially similar results which are not shown for the sake of brevity, but are available from the authors upon request.

9. Results are similar when using other strategies in order to weight product level tariffs and build firm-level tariffs, such as exploiting the products’ weight in the firm total production in the previous year. Results are available upon request.

10. Tariff data from WITS–TRAINS database are available at HS product level. In order to match tariff data with firm-product level data, we exploit the HS/CPA correspondence table, retrieved from Eurostat Ramon website, and we compute average tariffs for each CPA code.

11. When we cluster standard errors at the firm level, our results are not affected at all.

12. Furthermore, these findings are robust to the substitution of the continuous policy measures with dummy indicators for each quartile of the firm-level simple average tariff distribution (Lileeva and Trefler 2010). The only difference concerns the impact of export tariffs, as there is slight evidence that switching from high to low export costs increases firm diversification.

13. This evidence seems to be at odds with the evidence of learning by exporting found by part of literature (Wagner 2007). However, it could also suggest that the learning to export plays a major and more relevant role in Turkey. Before entering foreign markets, domestic firms get ready by adding new products, investing in tangible and intangible assets and engaging in productivity improvements. This anticipating effect can then drive the negative coefficient we find.

14. This effect may concern both established exporters and firms getting ready to enter export markets for the first time.