Adjustments in markups after a Free Trade Agreement: An analysis of Pakistani firms gaining increased access to China

ABSTRACT

Increased market access through trade liberalization can affect the markups, prices, and marginal costs of exporters. Understanding these dynamics is critical for firms and policymakers, particularly as they formulate export strategies. We examine the impact of China lowering tariffs on Pakistani products under the Pakistan–China Free Trade Agreement (FTA), which gave Pakistani exporters greater market access. Using disaggregated output and price data for textile manufacturers in Punjab, Pakistan, we estimate product-level markups and marginal costs using the methodology of De Loecker, Goldberg, Khandelwal, and Pavcnik (2016) [“Prices, Markups, and Trade Reform.” Economterica 84 (2): 445–510]. We then extend this to the firm level by using the methodology of De Loecker and Warzynsksi (2012) [“Markups and Firm-Level Export Status.” American Economic Review, 2437–2471]. We find that Pakistani firms exporting to China followed a dynamic pricing strategy by reducing prices to compete with global competitors in the Chinese market. We also find evidence of a decrease in marginal costs as a result of reductions in X-inefficiencies. But because Pakistan's exports to China are relatively homogeneous, the extent of quality differentiation and markup margins was limited. Finally, we find evidence of pro-competitive effects.

KEYWORDS:

• Firm-Level Productivity
• Trade Liberalization
• Free Trade Agreement
• Tariffs
• Markups
• X-inefficiency
• Dynamic Pricing
• Pro-competitive Effects

JEL CLASSIFICATIONS:

• D22
• D23
• F14
• F61
• L11

Disclosure statement

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

Notes

1 The products are classified into five segments as done by De Leocker (2011).

2 Input allocation across multiple products produced by a firm is hardly observed in any micro data set. Hence, many studies have made assumptions regarding this allocation based on the number of products (De Loecker 2011) and revenue shares (Foster et al. 2008).

3 Using this input-output relationship, a single product firm manufacturing motorcycles will use the same technology as a multiproduct firm manufacturing motorcycles and cars.

4 This expression can encompass both a value-added function and a gross output function. In the former case, only labor and capital enter the input set while in the former the input set in addition to labor and capital is a function other intermediate inputs e.g., materials.

5 This is the marginal cost since $\frac{\mathrm{\partial }{\mathcal{L}}_{\mathrm{it}}}{\mathrm{\partial }{Q}_{\mathrm{it}}}$= ${\lambda }_{\mathrm{it}}$

6 This expression for markup as a ratio of price over marginal cost is robust in various price (static) setting models and does not depend on a particular form of price competition amongst firms. However, it will depend on the specific nature of competition amongst firms. One restriction imposed is that prices are set period by period ruling out any cost adjustments of changing prices. Markups, however, will depend on the interaction amongst firms and the strategic interaction between them. We direct the reader towards the online appendix of De Leocker & Warzynski (2012) for discussion on some leading cases in this.

7 If one is further interested to back out the capital and labor coefficients based on the GNR (2020), the next step is to rely on partial differential equations for the production function and integrate them based on moment conditions on innovation in productivity which follows the Markov process. This last step then helps recover the capital and labor coefficients.

8 This means that our data set allows us observe the changes in the product mix for each firm at different points in time.

9 1 PKR equals to approximately $ 0.0044 as in 2022. The values reported in the table are current PKR.

10 Based on De Loecker’s (2011) classification, we divide the textile sector into five segments: (i) finishing (ii) spinning, (iii) interior, (iv) clothing, and (v) technical.

11 We multiply the coefficients by the average change in tariffs, which was 61.8% in our case to get the net impact. For example, the net impact of the FTA on the markups for firms exporting to China is 0.0882*0.618.

12 World Bank country classification is available at: https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending-groups