China’s competition and the export price strategies of developed countries

Abstract

This paper analyzes the impact of Chinese competition on developed countries’ export prices. The empirical application is on Italy, one of the main European manufacturing exporters with exports at high risk of competition from China. Our results show that, following China’s entry into the WTO, the price strategies of Italian firms has been affected. While in general the increasing Chinese export competition resulted in an upgrading of products exported, the impact has been different according to the sector and technological level. The incentives to upgrade have been stronger for low technology sectors, where competition is tougher and varieties of products sold lower. To highlight quality differentials, and isolate the effects on the different segments of the distribution of Italy’s export prices, we run quantile regressions. We find that are mainly those products sold at low prices to face a strong pressure to upgrade.

Keywords:

• China
• export price competition
• Italy
• heterogeneity

JEL Classifications:

• F-10
• F-14

Acknowledgements

We would like to thank an anonymous referee, Lionel Fontagné, Enrico Marvasi and participants to the 23rd Chinese Economic Association Conference at SOAS; the Italian Trade Study Group meeting in Catania; the 14th European Trade Study Group meeting at the Catholic University of Leuven; the Italian Society of Economists conference in Matera; the 2012 Columbia-Tsinghua Conference in International Economics at the Tsinghua University and the 10th C.Met Workshop at the University of Florence for their useful comments on previous drafts of the paper.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. The length of the lags is meant to vary according to the autocorrelation function.

2. Market shares have been computed at the product level as a simple ratio between the value of exports of any given exporter (China in this case) over the total exported value of the product for any couple of importer/year.

3. This is likely to allow us to control for the effects of quality differentials on the demand side.

4. Due to the difficulties in collecting data at such high levels of product disaggregation, most of the controls potentially influencing unit value could not be added to the empirical analysis.

5. Due to the presence of potential outliers, especially among the unit values (Table 1), we have run the analysis trimming out values of ‘p’ greater than the 99th percentile (i.e. greater that 13.8, or about US$10 million USD).

6. An additional econometric concern is the potential non-stationarity of the series used in the analysis. This is true as well for dynamic panel model, which assumes the series to be stationary (Blundell and Bond 1998). Although aware of the potential bias due to the presence of some non-stationary series in the data, we do not tackle this issue directly in this work considering the relatively short time dimension of the data as well as its highly unbalanced structure of the panel, preventing the adoption of some common test of non-stationarity.

7. We have run our model on additional levels of disaggregation, looking in particular at the most relevant groups of 2-digit industries in the consumer and the machinery groups. Results of these additional regressions, which are, in general, similar to those of Table 2, are not reported here for reasons of space, but are available on request.

8. According to Wooldridge (2010), although quantile regression analysis with fixed effects are feasible, their introduction in panels with large N and small T may bring the estimate to suffer from the incidental parameter problem.

9. Due to computational restrictions, we add sectoral fixed effects at the 2-digit level of the SITC classification rather than at the 6-digit level.