The gravity of surveys: nontariff barriers and the cost of traded goods

Transcription

1 The gravity of surveys: nontariff barriers and the cost of traded goods Joseph Francois (Johannes Kepler University Linz & CEPR) Miriam Manchin (University College London & Centro Studi Luca d'agliano) Hanna Norberg (Lund University) Frank van Tongeren (OECD) EAAE August 2011, Zürich

2 Summary Map survey-based market access rankings to trade flows and tariffs (gravity) Use estimated trade effect of NTBs with price elasticities and survey data to estimate feasible reduction in NTBrelated trade costs for G20 economies Feasible reductions: exploit intra-eu NTB preference margins Findings: Substantial efficiency gains possible, even under our realistic assumptions on feasible reductions Across all economies motor vehicles have highest NTB trade cost Imports into non-oecd G20 economies have substantially higher NTB costs than OECD members

3 Outline Market access rankings from business surveys Estimating trade costs from rankings Outcomes for industrial sectors and G20 economies Potential trade cost savings from more harmonization

4 What is wrong with business surveys? Business views on market access have little credibility in research: Selection bias: the squeaky wheel effect Often very specific issues Difficult to generalize across sectors and countries But business knows best where market access is difficult and where it is easy So: how to harness that knowledge for estimates of market access barriers?

5 The right questions yield better answers Simple questions (ECORYS 2009) Consider exporting to country X, keeping in mind your domestic market. If 0 represents a completely free trade environment, and 100 represents an entirely closed market due to NTMs, what value between would you use to describe the overall level of restrictiveness of country X to market to you export product (service) in this sector firms; bilateral NTM indexes (0-100) on home and foreign markets Good coverage EU and US, but also some info in third countries

6 Average indexes from survey NTB indexes Europe United States third intra-eea extra-eea intra- NAFTA extra- NAFTA country average ISIC 27,28 metals * ISIC textiles, clothing ISIC 20 wood products * ISIC 21,22 paper, pulp printing * ISIC 24,25 chemicals ISIC 30, 32 office machinery (electronics) * ISIC 29,31,33 other machinery * ISIC 34 motor vehicles * ISIC 35 other transport equipment

7 Convert NTM indexes to trade costs 2 steps: 1. Gravity equation with NTM index and tariffs on RHS 2. Back-out trade effect (=marginal effect) from change in index => estimate of ad valorem level of NTM

8 Gravity again From CES expenditure functions to import demand equations: Border price inclusive of trade cost Composite imports of k into country j Power of bilateral applied tariff Natural trade costs NTM cost multiplier And finally to this, with all RHS in logs: v ijs X k sk k ijs

9 Gravity - again v ijs X k sk k ijs X: tariff, ntmindex, distance etc. and importer and exporter fixed effects (take care of all country-pair unobservables) Can also compare with estimation of RTA effects (EU and NAFTA dummies) ZIP estimates because of many zero observations in bilateral trade flows Trade and tariff data from GTAP 7, distance etc from CEPII, ntmindex from survey After mapping all data: 142,486 obs, 99 countries, 10 industrial goods, year 2007

10 The trade effect of a change in (power of) tariff: dv tardt T 1 NTM trade effect per unit of tariff is: ntm tar dn ( 1 ) Note: Because ZIP estimation is V = exp(b ln(x)), need to take appropriate ln and exp transformations into account

11 50 years of harmonization have resulted in large reductions of intra-eu trade costs "average" intra-eu NTB margin total estimated EEA trade cost difference ISIC 27,28 metals ISIC textiles, clothing ISIC 20 wood products ISIC 21,22 paper, pulp printing ISIC 24,25 chemicals ISIC 30, 32 office machinery (electronics) ISIC 29,31,33 other machinery ISIC 34 motor vehicles ISIC 35 other transport equipment The intra-eu margin shows the % lower intra-eu trade cost relative to importing into EU from outside The EEA trade cost difference comes from gravity with EU dummy

12 50 years of harmonization have resulted in large reductions of intra-eu trade costs Trade Costs in Europe and NAFTA "average" intra-eu NTB margin total estimated EEA trade cost difference total estimated NAFTA trade cost difference ISIC 27,28 metals ISIC textiles, clothing ISIC 20 wood products ISIC 21,22 paper, pulp printing ISIC 24,25 chemicals ISIC 30, 32 office machinery (electronics) ISIC 29,31,33 other machinery ISIC 34 motor vehicles ISIC 35 other transport equipment The intra-eu margin shows the % lower intra-eu trade cost relative to importing into EU from outside The EEA and NAFTA trade cost difference comes from gravity with RTA dummies

13 Feasible reductions Total cost of good shipped 1+α 1+α Intra-EU NTM cost NTM-free price of good Reduction to zero infeasible (and does not maximize welfare): down to intra-eu level (1+α ) should be attainable: Efficiency gain: (α-α )/(1+ α)

14 Feasible NTM trade cost reductions Reduction to zero is not feasible (and would not maximize welfare). Take EU as benchmark: what would be cost saving if countries would move to EU level?

15 Feasible NTM trade cost reductions by G20 economies Highest potential efficiency gains in non-oecd economies.

16 Conclusions Ad valorem estimates of NTM trade costs from firms market access rankings are robust Allows theory- and evidence consistent import specification in market models to estimate welfare effects...but need to be cautious with defining correct shock: reduction to zero makes no sense: a certain amount of trade cost related to regulations is unavoidable Feasible reductions level the NTM-related trade costs across importers and imply efficiency gains