Standards Harmonization as Export Promotion

Transcription

1 Standards Harmonization as Export Promotion Marion Dovis University of Aix-Marseille Mélise Jaud The World Bank Exporters in MENA Workshop, The World Bank, December 11-12, 2013

2 Motivation Increasing role of non-tariff measures (NTMs) as tariffs continue to fall Increasing number of Regional Integration Agreements involve NTM harmonization, e.g. Morocco Effect of NTM harmonization on trade is complex Unlike tariff: not necessarily welfare-reducing or discriminatory Affects access to both foreign and domestic markets Limited evidence on the impact of NTMs harmonization on firms performance, mostly examine effect on foreign market access => We focus on impact on domestic market access and examine the effect of non-tariff import competition on developing countries domestic firms export performance (IM)

3 This Paper Q: How do NTM harmonizations affect import competition faced by domestic firms and their subsequent export performance? What s new? 1. Data : unique merged NTM-customs dataset for Morocco ( ) Measure of import competition at HS6 digit: NTM harmonization 2. Estimation strategy: DiD approach: we relate a change in non-tariff protection associated with NTM harmonization with change in domestic firms exports Exploit waves of NTM harmonization - with dates that vary across products and estimate the within firm-product effect of harmonization Focus on incumbent firms isolate import competition channel Identification issues: reverse causality, omitted variable bias

4 Underlying assumptions 1. Change in non-tariff measures have important impact on the competitive pressure faced by domestic firms Harmonization translated into implementation of stricter standard relative the one previously in place Exporters sell on the domestic market: over 60% of exporting manufacturing firms are active on the domestic market (census data), drop MNC (ZF) 2. Incumbent firms are already compliant. Firm that export a given product to harmonized destination markets throughout the sample period: 75% of incumbent transactions are with HIC income Not affected by compliance cost of stricter domestic standard Export cost unaffected Change in the non-tariff protection only affects incumbent firms export performance through changes in level of competition they face on the domestic market. 3. Level of income: proxy for countries harmonization status

5 Summary of Results 1. Harmonization of domestic NTM affect the export performance of domestic firms in import-competing industries through changes in the level of competition they face. Increased exports of incumbent firms and reduced their export price (MC production) At the extensive margin expansion of portfolio towards harmonized products 2. Effect vary with destination demand for harmonized product: Standard harmonization only confers exporters with a competitive edge in high income destination markets where demand for harmonized products is high Similarly at the extensive margin 3. Effect is sustained over time

6 Existing Literature At aggregate level: Positive effect of standard harmonization on trade flows lack of harmonization of standards and technical regulations is detrimental to trade (Otsuki, Wilson and Sewadeh, 2001; Gebrehiwet, Ngqangweni and Kirsten, 2007; Fontagné et al., 2005) Harmonization of domestic food standards on international ones had a positive and significant impact on exports (Mangelsdorf, Portugal-Perez and Wilson, 2012) Harmonization and market structure Chen and Mattoo (2004) harmonization on regional standards improves market access for out-of-bloc exporters from industrial countries, while reducing it if they were from developing countries. Harmonization of developing countries on developed countries standards as part of regional agreements reinforce hub-and-spoke trade structures at the expense of South-South trade (Disdier et al., 2012). At the firm-level Reyes (2012) impact of EU harmonization on US firms export performance: positive impact on firms ability to enter the EU market crowding out of developing countries producers. Augier, Cadot and Dovis (2013) harmonization of Moroccan domestic standards with international ones: positive impact of firms profitability and productivity. 6

7 Theory What should we expect I We formalize the intuition on harmonization s effects in a Melitz model following Demidova and Rodriguez-Clare (2011) : Model set up Two countries, home (i=1, Morocco) and foreign (i=2) τ 12 is an iceberg-cost type for firm located in 1 to sell in 2. Only difference: we model the effect of stricter standard as a rise in cost of selling in home market, that is a rise in τ 11, and a rise in τ 21 for non-compliant partners Open Economy equilibrium: Standards harmonization affect the level of competition on the domestic market through the change in relative trade cost τ21 /τ11 7

8 Theory What should we expect II 2 competing effects on the domestic market 1. Limit competition from low income producers: the standard is tougher to meet for foreign producers than for domestic o nes, relative cost τ21 /τ11 rises (iii) Exit of firms from export: raises the MPC of domestic exporters on the foreign market (iv) Higher domestic profits: by reducing the domestic real wage 2. Increase competition from High-income producers the trade cost of foreign exporters remains unchanged, τ11 rises so relative cost τ21 /τ11 shrinks (iii) Entry of Firms on export: reduces MPC of domestic exporters on the foreign market (iv) Lower domestic profits: by raising the domestic real wage 8

9 Theory What should we expect III Testable Predictions Intensive Margin : focus on incumbent firms less expensive to service domestic market relative to (i) low-income producers, (ii) domestic firms previously non compliant and (iii) High-income producers unambiguous positive effect on market shares : Raise Exports unambiguous negative effect on price Extensive Margin : ambiguous effect Effect depend on destination demand for harmonized product 9

10 Data 1. Moroccan customs: All annual export transaction flows between HS 10-digits level, in a four-dimensions panel: firm-year-product-partner (No data on firms characteristics) 2. NTM data: New database from World Bank National inventories of all trade regulations For each HS6 products whether a NTM is in place codes and whether it is harmonized with international regulations => Merged customs and NTM data at the product HS 6- digit level 10

11 NTM Data I Harmonization took place in waves: 2003, 2006 and 2012 NTM harmonized with an international standard represent for 14% of NTMs adopted in the Moroccan economy over the period Year # NTM # NTM Type NTM domestic harmonized A B D E F P

12 Customs-NTM Data, In aggregate and at the firm level : harmonized products represent a growing share in the number of product exported and the share of exports. Year N products Total export N firms Average firm N products Count Harmonize d product (%) Value (million USD) Harmonize d product (%) Count Harmoniz ed product (%) N destination s Count Harmonized product (%) Value (thousan d USD) Export Harmonized product (%) % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % Note: Harmonized products are products that underwent harmonization. Firms exporting harmonized products are defined as firms exporting at least one products affected by harmonization. 12

13 Descriptive Statistics Panel a: Harmonized vs. Non-Harmonized Products, Average Firm Export in USD Average Number of Firms Average firm exports value per product Panel b: Harmonized Products, Before and After Harmonization Harmonization time 13

14 Empirical Strategy Difference-in-Difference equation: YY ffffffff = αα + ββ 1 NNNNMM pppp 1 haaaaaa + ββ 2 NNNNMM pppp 1 haaaaaa IIIIIIIIIIIIIIIIII ffff + ββ 3 NNTTMM pppp 1 mmmmmm + γγ XX ffffffff + φφ ffff + φφ dd + φφ tt + ωω ffffffff (1) Estimate effect of product-specific treatment on firms export performance across destinations-years, with firm-product and time effects Investigate import-competition channel : focus on incumbent firms export a given product to harmonized destination markets throughout the sample period Not affected by compliance cost of stricter domestic standard Change in the non-tariff protection only affects incumbent firms export performance through changes in level of competition they face on the domestic market. Yfdpt = log export quantity in kg or log export UV Standards Harmonization As Export Promotion 14

15 Main Result: Standards Harmonization, Import Protection and Firm Exports Export quantity Unit Value (1) (2) (3) (4) NTM harm t * (0.0377) (0.0190) NTM harm t-1*incumbent *** * (0.0606) (0.0311) NTM harm t-1*nonincumbent (0.0563) (0.0204) NTM mor (0.0395) (0.0394) (0.0216) (0.0216) Nbexporters *** *** *** *** (0.0105) (0.0105) (0.0031) (0.0031) Tariff t *** *** *** *** (0.0099) (0.0099) (0.0039) (0.0039) Import *** *** (0.0183) (0.0183) (0.0072) (0.0072) Constant *** *** *** *** (0.1739) (0.1738) (0.1129) (0.1129) product-firm Fixed Effects year destination Observations R-squared Robust standard errors in parentheses clustered by product-year. * significant at 10%; ** significant at 5%; ***significant at 1%. 15

16 Destination Income Effect: Proxy for Demand for Harmonized Products 16

17 Standards Harmonization Promotes Existing Firms Entry into Harmonized Products Export YY ffffffff = αα + ββ 1 NNNNMM pppp 1 haaaaaa + ββ 2 NNTTMM pppp 1 mmoooo + γγ XX ffffffff + φφ ffff + φφ dd + φφ tt + εε ffffffff (2) All Entry firm-product High Non-high income income All Product creation High income Non-high income (1) (2) (3) (4) (5) (6) NTM harm t *** *** (0.0185) (0.0207) (0.0285) (0.0112) (0.0131) (0.0147) NTM mor (0.0157) (0.0190) (0.0193) (0.0113) (0.0143) (0.0123) Nbexporters *** *** *** *** (0.0014) (0.0015) (0.0043) (0.0011) (0.0013) (0.0026) Tariff t *** ** (0.0015) (0.0021) (0.0022) (0.0011) (0.0016) (0.0016) Import ** * (0.0047) (0.0045) (0.0129) (0.0036) (0.0037) (0.0074) Constant ** * ** (0.0239) (0.0178) (0.0445) (0.0184) (0.0138) (0.0424) product-firm Fixed Effects year destination Observations 289, ,657 40, , ,657 40,839 R-squared Robust standard errors in parentheses clustered by product-year. * significant at 10%; ** significant at 5%; ***significant at 1%. 17

18 Standards Harmonization and Spillover Effects Promotes Existing Firms Entry into non-harmonized Products Export Within Same Sector Product creation All High income Non-high income (1) (2) (3) NTM harm t *** *** NTM harmhs4 t * ** NTM mor Nbexporters *** *** Tariff t ** * Import ** * Constant ** product-firm Fixed Effects year destination Observations 289, ,657 40,839 R-squared Robust standard errors in parentheses clustered by product-year. * significant at 10%; ** significant at 5%; ***significant at 1%. 18

19 Standards Harmonization Effects Persist Over Time 19

20 Robustness Checks I : Intensive Margin Excluding top and bottom 1% of firms in our sample Unit value Export quantity High No high High All All income income income (1) (2) (3) (4) (5) (6) Excluding firms in top & bottom 1% (total export) No high income NTM harm t-1*incumbentfp ** ** * (0.0583) (0.0706) (0.0983) (0.0315) (0.0358) (0.0414) NTM harm t-1*no Incumbentfp (0.0555) (0.0647) (0.0957) (0.0205) (0.0233) (0.0416) Observations R-squared Robust standard errors in parentheses clustered by product-year. * significant at 10%; ** significant at 5%; ***significant at 1%. 20

21 Robustness Checks I : Extensive Margin Excluding top and bottom 1% of firms in our sample All Entry firm-product High income All Product creation High income Nonhigh income Nonhigh income (1) (2) (3) (4) (5) (6) All sectors NTM harm t *** *** (0.0185) (0.0209) (0.0288) (0.0113) (0.0131) (0.0151) Observations R-squared Observations R-squared Robust standard errors in parentheses clustered by product-year. * significant at 10%; ** significant at 5%; ***significant at 1%. 21

22 Robustness Checks III : Placebo Mean Mean t- Replication Mean S.E coefficient stat number Dependent Variable (1) (2) (3) (4) Export quantity Baseline Incumbent Incumbent and high income sample Export growth of quantity Baseline Incumbent Incumbent and high income sample Unit value Baseline Incumbent Incumbent and high income sample Product creation Baseline high income sample No high income sample

23 Concluding Remarks Estimate the within firm-product import-competition effect of standard harmonization on firms exports 1. Harmonization of domestic NTM affect the export performance of domestic firms in import-competing industries through changes in the level of competition they face. Increased exports of incumbent firms and reduced their marginal production cost At the extensive margin expansion of portfolio towards harmonized products 2. Effect vary with destination demand for harmonized product: Standard harmonization only confers exporters with a competitive edge in high income destination markets where demand for harmonized products is high Similarly at the extensive margin. 3. Effect is sustained over time Harmonization of developing countries domestic standards with stricter international regulations may serve as a policy instrument to promote domestic firms exports.

24 Next Steps 1. Heterogeneity of the effect across firm size 2. Heterogeneity of the effect across sectors 3. Disentangle the two competing effects: distinguishing sectors with high penetration of low-income competitors versus sectors with high income competitors Market protection Procompetitive effect 24

25 Thank You 25

26 Motivation II What are we talking about? NTMs are requirements on product characteristics labeling, price control and quantity, and conformity assessment, that define conditions for access to a given market NTMs harmonization btw 2 countries adopt a common standard which may be one of theirs or a third one (regional, i.e. new, or existing international) Morocco: vigorous effort to modernize and harmonize with international standards - largely mandated under the Association Agreement with the EU and WTO agreements