Impact of technical barriers to trade measures on innovation – evidence from chinese manufacturing firms

ABSTRACT

This study continues the discussion of the relationship between globalization and innovation. Unlike existing studies, this paper shifts attention to the impact of changes in the trade environment caused by Technical Barriers to Trade (TBT) measures on innovation. With Chinese Manufacturing firms’ micro panel dataset and using the number of firm patent applications as a proxy variable to measure innovation, we empirically test the effect of TBT measures on this innovation. This study further distinguishes the heterogeneous effects of the total TBT measures and highly restrictive measures among them. The main estimation results indicate that the total amount of TBT measures imposed by trading partners does not significantly affect the number of patent applications by Chinese exporting firms. However, highly restrictive TBT measures demonstrate a significant negative effect. This once again proves that not all TBT measures are harmful. In addition, we obtained more interesting results through further heterogeneity analyses.

KEYWORDS:

• Technical barriers to trade measures
• innovation
• Chinese manufacturing firms

JEL Classifications:

• F13
• O31

Data availability statement

The data that support the findings of this study are available Social Science Data Integrated Service Platform. Restrictions apply to the availability of these data, which were used under license for this study. Data are available www.ppmandata.cn with the permission of Social Science Data Integrated Service Platform.

The other data that support the findings of this study are openly available in UNCTAD TRAINS Database and wiiw Database at https://wits.worldbank.org and https://wiiw.ac.at/wiiw-ntm-data-ds-2.html.

Disclosure statement

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

Notes

1 According to the WTO Integrated Trade Intelligence Portal (I-TIP) database, as of 2022, the number of SPS and TBT measures notified by WTO members accounts for approximately 36% and 53% of the total number of NTMs, respectively.

2 We do not use the number of grants because patents often take a long time from application to grant. However, when a firm applies for a patent, it means it has already innovated. Therefore, the number of applications can better reflect the degree of innovation of the firm in the current period.

3 We only focus on TBT measures since our empirical samples are all manufacturing firms, but SPS is concentrated in the agricultural and food products sector.

4 Special Trade Concerns refer to concerns raised by WTO members and observers regarding technical NTMs that are discriminatory, overly restrictive, and trade protectionist. This can be understood as a grievance mechanism. Therefore, regulatory measures that trigger STC can be considered highly restrictive.

5 According to Ghodsi et al. (2017), China ranked first among the countries most frequently targeted by NTMs during the period 1995 to 2014.

6 For a detailed introduction to the PPML estimator, please refer to Section 4.

7 According to Fontagné et al. (2015), UNCTAD (2018), and Shepherd and Peters (2020), additional compliance costs mainly include (i) direct compliance costs, such as the cost of adjusting production processes or products; (ii) indirect costs, such as testing required to demonstrate compliance and border delays; (ii) increased procurement costs resulting from the use of more expensive inputs.

8 Shepherd and Peters (2020) pointed out that the increase in trade costs caused by NTMs is approximately three times that of tariffs.

9 The consistent correspondence table between HS versions comes from the World Integrated Trade Solution (WITS).

10 Exchange rate data comes from the World Bank’s World Development Indicators.

11 Where the inforce date is missing, we use the initiation date instead. In general, most NTMs are initiated and enforced in the same year.

12 For detailed steps, see Ghodsi et al. (2017).

13 These firms belong to different legal systems.

14 During the period 1998-2006, the firm-level production data set includes all SOEs and non-SOEs with sales of 5 million yuan and above. However, during the period 2007-2011, the data set records all industrial firms with sales of 5 million yuan and above. In 2011 and after, the data set only covers industrial firms with 20 million yuan and above sales. Therefore, to unify the statistical caliber, we deleted firms whose sales were always less than 20 million yuan.

15 Appendix Table A1 lists the name of the manufacturing sector corresponding to each CIC 2-digit code.

16 These variables include total assets, number of employees, gross value of industrial output, total fixed assets, and sales. As a robustness check, we also performed an estimation using the full sample.

17 Since some critical firm-level information required for empirical analysis was not reported in 2010, we had to abandon 2010.

18 We also logged the dependent variable after adding one to perform regular Ordinary Least Squares (OLS) estimation and obtained estimates similar to the baseline results. Regular OLS estimation results are available upon request.

19 We also attempted to regress the two together and obtained very similar results. The estimation results are available upon request.

20 Firm age is likely to be exogenous to the level of innovation.

21 According to the World Integrated Trade Solution (WITS), the effectively applied tariff is defined as the lowest available tariff. If a preferential tariff exists it is used as effectively applied tariffs, otherwise the MFN applied tariff is used.

22 For example, possible regional trends include different innovation support policies and differences in science and technology development levels.

23 Data for the CPI comes from the World Bank’s World Development Indicators.

24 The variables N.TBT and N.TBT_STC imposed by countries with different development levels and coverage indexes in Table 1 that are not introduced in this section are used for the subsequent heterogeneity analysis in Section 5.2 and the robustness check in Section 6.1, respectively.

25 The percentage change in the number of firms’ patent applications from an increase in the number of TBT-STC by one percentage point is calculated as follows: $[\exp (1\times \mathrm{coefficient})-1]\times 100$.

26 We divided the TBT-imposed countries in the sample according to the level of national development defined by World Economic Situation and Prospects 2014.

27 For a detailed classification of industry sectors, see Appendix Table A1.

28 In our sample, the export revenue of foreign-owned firms accounts for about 34% of sales, while that of domestic firms is about 24%.

29 Since introducing this fixed effect would cause us to lose more sample observations, it is not used in the main estimation.