Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions The Value of Deep Trade Agreements in the Presence of Pricing-to-Market Meredith A. Crowley, Lu Han, & Thomas Prayer World Bank Deep Trade Agreements Webinar November 2020
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Introduction • Prior research using the universe of international trade transactions for the UK (2010-2017) and China (2000-2014) has found that firms that export to multiple foreign destinations employ different pricing strategies. • Pricing-to-market is correlated with observables and is more prevalent for: • highly differentiated products, • consumer versus intermediate goods, • goods exported by foreign-invested firms (China), • goods invoiced in the local currency of the destination (UK). • Our question: Do preferential trade agreements lead to greater market integration, more intense competition, and less market power for exporting firms?
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Our approach Using product-level exports from 640k firms located in 13 emerging and low-income countries to 165 destinations, we examine 257 deep trade agreements to evaluate how • third-country competitive pressures (measured as PTAs and preferential tariffs) and • specific (bilateral) trade agreement provisions impact • firm-level export sales, • prices, and • destination-specific markups.
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Literature • Gravity Anderson & vanWincoop 2003, Feenstra 2004, Redding &Venables 2004, Baier & Bergstrand 2007, Head & Mayer2014; • Price and markup responses to trade policy Konings & Vandenbussche 2005, Bown & Crowley 2006, Amiti & Konings 2007, Pierce 2011, deLoecker, Goldberg, Khandelwal, & Pavcnik 2016; • Third-country competition and trade-policy spillovers Chang & Winters 2002, Bown & Crowley 2007, Romalis 2007, Estevadeordal, Freund, & Ornelas, 2008, Corsetti, Crowley, Han, and Song 2018.
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Firms’ product-level exports from 13 origin countries 27.5 million firm-product-origin-destination-year observations Albania 2004-2012 Guatemala 2005-2013 Senegal 2000-2012 Burkina Faso 2005-2012 Jordan 2003-2012 Uruguay 2001-2012 Bulgaria 2001-2006 Mexico 2000-2012 Yemen 2008-2012 China 2000-2006 Malawi 2006-2012 Egypt 2005-2016 Peru 1993-2013 HS06 product-level data on destination trade policy against origin AND destination trade policy against origin’s competitor countries.
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Structural gravity equation applied to firm-level data We start with a model of trade flows as a function of preferential trade agreements (PTAs) and a set of fixed effects. ln ( v fodit ) = β 1 ∗ pta odt + δ foit + δ dit + ( δ od ) + ǫ fodit (1) where • f , o , d , i , t represent firm, origin country, destination country, product and year respectively. • δ foit : firm-origin-product-year fixed effects • δ dit : destination-product-year fixed effects • δ od : origin-destination fixed effects
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Structural gravity equation extended to include a measure to capture competition from third countries We then introduce a measure to capture the share of country o ’s competitors that have a PTA with destination d . ln ( v fodit ) = β 1 ∗ pta odt + β 2 ∗ competitor pta ( − o ) dit + (2) + δ foit + δ dit + ( δ od ) + ǫ fodit where • f , o , d , i , t represent firm, origin country, destination country, product and year respectively.
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Third-country competition effects Calculating the Proportion of Competitor Countries with Access to a PTA • Origin (bottom) exports to destination (center) • Firms in five competitor countries also export to destination • Countries 1-3 have PTAs with destination • Countries 4-5 do NOT • Product-level import-wt’d share of competitor countries with PTA = 0.6
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Structural gravity equation extended to include measures of tariffs and trade agreement provisions We extend the specification to include: • tariffs (on origin and on third-country competitors) and • PTA provisions (for origin and third-country competitors). ln ( v fodit ) = β 1 ∗ pta odt + β 2 ∗ competitor pta ( − o ) dit + β 3 ∗ ln (1 + bilateral τ odit ) + β 4 ∗ ln (1 + competitor τ ( − o ) dit ) + β 5 ∗ provision odt + β 6 ∗ competitor provision ( − o ) dit + δ foit + δ dit + ( δ od ) + ǫ fodit The analysis can be applied to any (binary) provision in the DTA database.
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Third-country competition effects Calculating the Average Tariff Faced by Competitors • Origin (bottom) exports to destination (center) • Firms in five competitor countries also export to destination • Country 4 faces a 0% tariff in destination • Countries 1-3 and 5 face non-zero tariffs • Product-level import-wt’d average tariff faced by competitor countries = 4.5%
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Main findings: PTAs • Direct effects of PTAs: • Export sales increase 40-52% • Prices fall 2-5% • Markups fall 4% PTAs have pro-competitive effects. • Indirect, third-country PTA effects: • Export sales fall 10% when 10% of competitors have a PTA • Prices increase 0.4-0.8% when competitors’ tariffs increase 1% • Markups of highly differentiated goods increase 1.4% when competitors’ tariff increase 1%
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Deep Trade Agreements PTAs and third-country competition effects on average export value from origin Value Value Value If origin and destination (1) (2) (3) have a PTA ⇒ avg exports PTA odt 0.42*** 0.34*** -0.04*** from origin ↑ 40% Competitors’ Avg PTA ( − o ) dit -1.12*** -0.36*** If 10% of origin’s competitors in destination Tariff odit -4.02*** -3.73*** -1.25*** have a PTA ⇒ avg exports Competitors’ Avg Tariff ( − o ) dit 4.83*** 3.71*** from origin ↓ 10.6% Observations 16, 338,526 15,543,005 15,542,843 A 1% increase in the avg Fixed Effects tariff against competitors in Origin-firm-product-year � � � destination ⇒ avg exports Destination-product-year � � � Origin-destination � from origin ↑ 4.8%
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Deep Trade Agreements PTAs and third-country competition effects on average export value from origin Inclusion of O-D fixed effects implies PTAs have a But trade policy against small negative effect on trade. third-country competitors remains important: Value Value Value (1) (2) (3) • If 10% of origin’s PTA odt 0.42*** 0.34*** -0.04*** competitors in Competitors’ Avg PTA ( − o ) dit -1.12*** -0.36*** destination have a PTA ⇒ avg exports Tariff odit -4.02*** -3.73*** -1.25*** from origin ↓ 3.5% Competitors’ Avg Tariff ( − o ) dit 4.83*** 3.71*** • A 1% increase in the Observations 16, 338,526 15,543,005 15,542,843 avg tariff against Fixed Effects competitors in Origin-firm-product-year � � � Destination-product-year � � � destination ⇒ avg Origin-destination � exports from origin ↑ 3.7%
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Estimating the impact on prices (unit values) and markups ln ( p fodit ) = β 1 ∗ pta odt + β 2 ∗ competitor pta ( − o ) dit + β 3 ∗ ln (1 + bilateral τ odit ) + β 4 ∗ ln (1 + competitor τ ( − o ) dit ) + β 5 ∗ provision odt + β 6 ∗ competitor provision ( − o ) dit + δ dit + ( δ f & δ oit ) or ( δ od & δ oit ) or δ foit + ǫ fiodt We regress prices on destination-product-time fixed effects and: 1. firm and origin-product-time fixed effects, or 2. origin-destination and origin-product-time fixed effects, or 3. firm-origin-product-time fixed effects (to study markups).
Intro Data Empirical Strategy Results: PTAs Results: PTA Provisions Conclusions Deep Trade Agreements Pro-competitive effect of PTAs on prices and markups If origin and destination Prices Prices Markups have a PTA ⇒ avg prices (4) (5) (6) from origin ↓ 2-5% PTA odt -0.02*** -0.05** -0.04*** If origin and destination Competitors’ Avg PTA ( − o ) dit -0.02 -0.13*** 0.01 have a PTA ⇒ avg markups from origin ↓ 4% Tariff odit 0.00 0.24* -0.03 Prices are increasing in own Competitors’ Avg Tariff ( − o ) dit 0.45* 0.86** 0.25 tariff and in the avg tariff Observations 14,256,684 14,257,049 14,238,884 against competitors when Fixed Effects origin-destination FEs Origin-firm-product-year � included AND when sample Destination-product-year � � � Origin-destination � restricted to highly Origin-product-year � � differentiated goods (not Firm � shown).
Recommend
More recommend
