Quality, trade, and exchange rate pass-through

    Quality, Trade, and Exchange Rate Pass-Through Natalie Chen, Luciana Juvenal PII: DOI: Reference:

S0022-1996(16)30007-1 doi: 10.1016/j.jinteco.2016.02.003 INEC 2932

To appear in:

Journal of International Economics

Received date: Revised date: Accepted date:

16 July 2015 15 February 2016 17 February 2016

Please cite this article as: Chen, Natalie, Juvenal, Luciana, Quality, Trade, and Exchange Rate Pass-Through, Journal of International Economics (2016), doi: 10.1016/j.jinteco.2016.02.003

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

RI P

T

Quality, Trade, and Exchange Rate Pass-Through∗† Luciana Juvenal International Monetary Fund

SC

Natalie Chen University of Warwick, CAGE, CESifo, and CEPR

NU

February 24, 2016

MA

Abstract

CE

PT

ED

We investigate theoretically and empirically the effects of real exchange rate changes on the behavior of firms exporting multiple products with heterogeneous levels of quality. Our model, which features a demand elasticity that falls with quality, predicts more pricing-to-market and a smaller response of export volumes to a real depreciation for higher quality goods. We provide strong support for the model predictions using a unique data set of Argentinean firm-level wine export values and volumes between 2002 and 2009 combined with experts wine ratings to measure quality. The heterogeneity we find in the response of export prices and volumes to changes in exchange rates remains robust to alternative measures of quality, samples, and specifications.

AC

JEL Classification: F12, F14, F31 Keywords: Distribution costs, exchange rate pass-through, exports, heterogeneity, multi-product firms, pricing-to-market, quality, wine.

∗ We are grateful to the editor Charles Engel and to two anonymous referees for many valuable comments. We thank the Centre for Competitive Advantage in the Global Economy (CAGE) at the University of Warwick, the Federal Reserve Bank of St Louis, and the International Monetary Fund for financial support. We thank Kym Anderson, Nicolas Berman, Illenin Kondo, Logan Lewis, Thierry Mayer, Guillermo Noguera, Dennis Novy, Herakles Polemarchakis, Lucio Sarno, C´edric Tille, Olga Timoshenko, Zhihong Yu, our discussants Raphael Auer, Julian di Giovanni, Fadi Hassan, Philippe Martin, Ferdinando Monte, Tim Schmidt-Eisenlohr, and Charles van Marrewijk, and participants at the Banque de France and Sciences Po Workshop on Recent Developments in Exchange Rate Economics, CAGE Research Meetings 2013, CEPR ESSIM 2014, CEPR Exchange Rates and External Adjustment Conference Zurich 2014, CEPR Tenth Annual Workshop on Macroeconomics of Global Interdependence Dublin 2015, EITI Phuket 2014, ETSG Birmingham 2013, Federal Reserve Bank of Atlanta and NYU’s Stern School of Business Second Annual Workshop on International Economics, INFER Conference on Regions, Firms, and FDI Ghent 2014, Midwest International Trade Meetings 2013, Spring 2015 Midwest Macro Meetings, Tri-Continental International Trade Policy Conference Beijing 2014, DC Trade Study Group, Geneva Trade and Development Workshop, Inter-American Development Bank, International Monetary Fund, London School of Economics, Lund, Nottingham, Utrecht School of Economics, and Warwick. Brett Fawley provided excellent research assistance in the early stages of this project. The views expressed are those of the authors and do not necessarily reflect official positions of the International Monetary Fund. An online appendix that accompanies this paper can be found on the authors’ websites. † Natalie Chen, Department of Economics, University of Warwick, Coventry CV4 7AL, UK. E-mail: [email protected] (corresponding author) and Luciana Juvenal, International Monetary Fund. E-mail: [email protected].

ACCEPTED MANUSCRIPT 1

Introduction

SC

RI P

T

Exchange rate fluctuations have small effects on the prices of internationally traded goods. Indeed, empirical research typically finds that the pass-through of exchange rate changes to domestic prices is incomplete, leading to deviations from the Law of One Price.1 Possible explanations for partial passthrough include short run nominal rigidities combined with pricing in the currency of the destination market (Engel, 2002; Gopinath and Itskhoki, 2010; Gopinath, Itskhoki, and Rigobon, 2010; Gopinath and Rigobon, 2008), pricing-to-market strategies whereby exporting firms differentially adjust their markups across destinations depending on exchange rate changes (Atkeson and Burstein, 2008; Knetter, 1989, 1993), or the presence of local distribution costs in the importing economy (Burstein, Neves, and Rebelo, 2003; Corsetti and Dedola, 2005).2

ED

MA

NU

Thanks to the increasing availability of highly disaggregated firm- and product-level trade data, a strand of the literature has started to investigate the heterogeneous pricing response of exporters to exchange rate changes.3 Amiti, Itskhoki, and Konings (2014) find that Belgian exporters with high import shares and high export market shares have a lower exchange rate pass-through. Berman, Martin, and Mayer (2012) show that highly productive French exporters change significantly more their export prices in response to real exchange rate changes, leading to lower pass-through. Chatterjee, DixCarneiro, and Vichyanond (2013) focus on multi-product Brazilian exporters and find that within firms, pricing-to-market is stronger for the products firms are most efficient at producing.

AC

CE

PT

Evidence on the role of product-level characteristics in explaining heterogeneous pass-through remains scarce, however. In this paper, we explore how the heterogeneous pricing-to-market behavior of exporters, which leads to incomplete pass-through, can be explained by the quality of exported goods. We model theoretically the effects of real exchange rate changes on the pricing decisions of multi-product firms that are heterogeneous in the quality of the goods they export, and empirically investigate how such heterogeneity impacts exchange rate pass-through for export prices. Consistent with the predictions of the model, our empirical analysis shows that pass-through decreases with the quality of exported goods. Assessing the role of quality in explaining pass-through is a challenge as quality is generally unobserved. To address this issue we focus on the wine industry, which is an agriculture-based manufacturing sector producing differentiated products, and combine a unique data set of Argentinean firm-level destination-specific export values and volumes of highly disaggregated wine products with experts wine ratings as a directly observable measure of quality.4 The first contribution of the paper is to present a theoretical model to guide our empirical specifications. Building on Berman et al. (2012) and Chatterjee et al. (2013), we extend the model of Corsetti and Dedola (2005) by allowing firms to export multiple products with heterogeneous levels of quality. 1

For a survey of the literature, see Burstein and Gopinath (2014) and Goldberg and Knetter (1997). Also, Alessandria and Kaboski (2011) emphasize the role of search frictions as a source of incomplete pass-through. Nakamura and Steinsson (2012) argue that price rigidity and product replacements lead aggregate import and export price indices to appear smoother than they actually are, biasing exchange rate pass-through estimates. 3 Many papers examine the response of import prices (which include transportation costs) or consumer prices (which further include distribution and retail costs) to changes in currency values (e.g., Campa and Goldberg, 2005, 2010; Gopinath and Rigobon, 2008). For earlier evidence from the perspective of exporters, see Goldberg and Knetter (1997) and Knetter (1989, 1993). For more recent evidence at the firm-level, see Campos (2010), Fitzgerald and Haller (2014), and Li, Ma, and Xu (2015). 4 Other papers that focus on specific industries include Auer, Chaney, and Saur´e (2014) and Goldberg and Verboven (2001) for cars, Hellerstein (2008) for beer, and Nakamura and Zerom (2010) for coffee. 2

1

ACCEPTED MANUSCRIPT

SC

RI P

T

In the presence of additive (per unit) local distribution costs paid in the currency of the importing country, which are assumed to be higher for higher quality goods, the model shows that the demand elasticity perceived by firms falls with a real depreciation and with quality. In response to a change in the real exchange rate, exporters therefore change their prices (in their own currency) more, and their export volumes less, for higher quality goods. As a result, pass-through decreases with quality. Once we allow higher income countries to have a stronger preference for higher quality goods, as the evidence from the trade literature tends to suggest (e.g., Crin` o and Epifani, 2012; Hallak, 2006), the heterogeneous response of prices and quantities to exchange rate changes is predicted to be stronger for exports to higher income destination countries.

PT

ED

MA

NU

In contrast to models where quality is only a demand shifter (e.g., Melitz, 2003; Khandelwal, Schott, and Wei, 2013), our extension to Corsetti and Dedola (2005) with local distribution costs predicts that the demand elasticity perceived by exporters falls with quality. Other models with endogenous and variable markups deliver the prediction that quality affects the perceived demand curve, and that passthrough decreases with quality: Atkeson and Burstein (2008) with imperfect competition ` a la Cournot where high performance firms have high market shares; Melitz and Ottaviano (2008) with fixed costs of entry and non-CES preferences where high performance firms have a lower price elasticity of demand; and Auer, Chaney, and Saur´e (2014) with monopolistic competition and non-homothetic preferences where high market share firms perceive a lower demand elasticity. Overall, by showing that passthrough decreases with quality, our empirical analysis validates all these classes of models, but we also provide evidence, consistent with our Corsetti and Dedola (2005) framework, that distribution costs matter in explaining our findings.

AC

CE

The second, and main, contribution of the paper is to bring the predictions of our model to the data. The firm-level trade data we rely on are from the Argentinean customs and provide, for each export flow between 2002 and 2009, the name of the exporting firm, the country of destination, the transaction date, the Free on Board (FOB) value of exports (in US dollars), and the volume (in liters) of each wine exported, where a wine is defined according to its name, grape (Chardonnay, Malbec, etc.), type (white, red, or ros´e), and vintage year. As we do not directly observe prices, we compute FOB export unit values as a proxy for export prices at the firm-product-destination level, using data on the value and volume exported. To assess the quality of wines, we rely on two well-known experts wine ratings, the Wine Spectator and Robert Parker. Our approach to measuring quality is similar to Crozet, Head, and Mayer (2012) who match French firm-level exports of Champagne with experts quality ratings to investigate the relationship between quality and trade. Our data are well-suited for identifying heterogeneous pass-through due to differences in product quality. First, as the level of disaggregation of the customs data is unique, we can define a “product” in a much more precise way compared to papers that rely on trade classifications such as the Combined Nomenclature (CN) or the Harmonized System (HS) to identify products (e.g., Amiti et al., 2014; Auer and Chaney, 2009; Berman et al., 2012; Chatterjee et al., 2013). Second, as the quality scores from the Wine Spectator and Parker rating systems are assigned to each wine according to its name, grape, type, and vintage year, the trade and quality data sets can directly be merged with each other. As a result, when using the Wine Spectator ratings that have the largest coverage of Argentinean wines, our sample includes 209 multi-product firms exporting 6,720 different wines with heterogeneous levels of quality (this contrasts with Crozet et al., 2012, who cannot distinguish between the different 2

ACCEPTED MANUSCRIPT

NU

SC

RI P

T

varieties sold by each firm, and therefore assume that each firm exports one type of Champagne only). Aggregating our data at the 12-digit HS-level would reduce our sample size by a factor of six, which would in turn significantly lower the within-firm variation in the quality of exported wines as the 209 exporters would only be selling at most five different “products.” Third, the level of disaggregation of our data ensures that compositional or quality changes do not affect movements in unit values. For instance, Feenstra and Romalis (2014) demonstrate that the variation in trade unit values defined at the 4-digit Standard International Trade Classification (SITC) level is mostly attributable to differences in product quality. Fourth, thanks to the granularity of our data, our pass-through estimates are not affected by the “product replacement bias” put in evidence by Nakamura and Steinsson (2012), which may be a problem when using aggregate data (also, see Gagnon, Mandel, and Vigfusson, 2014). Finally, export values are FOB and therefore measure the revenue received by exporters at the border, excluding nontradable components such as transportation costs, tariffs, or distribution and retail costs in the importing country.

AC

CE

PT

ED

MA

To the best of our knowledge, our paper is the first to exploit such highly disaggregated productlevel exports data to investigate empirically the pricing strategies of exporters in response to real exchange rate fluctuations.5 We find that pass-through for export prices is incomplete, but large: in our baseline regression, in response to a ten percent change in the real exchange rate, exporters change their export prices (in Argentinean pesos) by 1.9 percent, therefore pricing-to-market is low and passthrough is large at 81 percent. Higher quality wines are more expensive and are exported at a higher price. Most interestingly, we show that the response of export prices to real exchange rate changes increases with the quality of exported wines, or in other words pricing-to-market rises with quality (i.e., pass-through decreases with quality). On average, pricing-to-market is 10.3 percent higher for a wine at the 95th than at the 5th percentile of the quality distribution. This heterogeneity in the response of export prices to exchange rate changes remains robust to different measures of quality, samples, and specifications. Consistent with our model, we also find that firms price-to-market more, and by more for higher quality wines, when local distribution costs are high. Export volumes increase in response to a real depreciation, but by less for higher quality goods. Finally, we find that the effect of quality on the responses of export prices and volumes to changes in exchange rates is larger for exports to higher income countries. One concern with our estimations is the potential endogeneity of quality in explaining unit values and export volumes. Although both the Wine Spectator and Parker rating systems are based on blind tastings where the price of each wine is unknown, the tasters are told the region of origin or the vintage year of each wine. As a result, we cannot rule out that the basic information that the experts hold about each wine, in combination with their own subjective preferences, affects in a way or the other their scores, leading to an endogeneity bias. In addition, as wine producers are likely to set their prices depending on the scores that are awarded to their wines, measurement error in the quality ratings is another potential source of endogeneity. To overcome this issue, we use instruments for quality based on geography and weather-related factors, including the total amount of rainfall and the average temperature during the growing season for each province where the grapes are grown, as well as the altitude of each of the growing regions. Our findings remain robust to the instrumentation of quality. 5 Papers using similarly detailed product-level data usually observe retail or wholesale prices rather than trade prices (e.g., Hellerstein, 2008; Nakamura and Zerom, 2010).

3

ACCEPTED MANUSCRIPT

SC

RI P

T

We also provide extensions to our benchmark specifications. First, when allowing for nonlinearities in the effect of quality, the elasticity of unit values to the exchange rate is shown to be much larger for the top quality wines, a finding which can be reconciled with the expectation that higher quality goods have higher (per unit) distribution costs. Second, pricing-to-market increases with firm-level productivity (Berman et al., 2012; Chatterjee et al., 2013). Third, higher quality firms price-to-market more than others. Finally, firms price-to-market more in the countries where they own a larger share of the market (Amiti et al., 2014; Atkeson and Burstein, 2008). Interestingly, quality matters beyond the effect of market shares, and the contribution of quality in explaining pass-through is larger than that of market shares.

PT

ED

MA

NU

Although our analysis focuses on a specific sector in a single country, evidence suggests that quality heterogeneity matters more generally, and that differences in the quality of traded goods across countries and industries are large (Feenstra and Romalis, 2014). Therefore, under the assumption that our results extend beyond the Argentinean wine industry, they carry macroeconomic implications. First, they provide evidence that quality and per capita income, by increasing pricing-to-market, play a key role in generating deviations from the Law of One Price, and therefore matter to explain the degree of international market segmentation. Second, as they improve our understanding of how firms set their prices when exporting abroad, and on how they react to exchange rate changes, they have implications for aggregate pass-through, and therefore for the evolution of the trade balance and the choice of exchange rate policy (Knetter, 1989). In particular, as incomplete pass-through determines how depreciations can stimulate an economy by substituting foreign by domestic goods, our results suggest that, by weakening their currency, lower income countries that specialize in lower quality goods are likely to improve their trade balance by more compared to higher income countries that produce, on average, higher quality goods.

AC

CE

Our paper belongs to three strands of the literature. The first one is the vast literature on incomplete pass-through and pricing-to-market. These papers usually find a low degree of pass-through into import prices (e.g., Campa and Goldberg, 2005; Gopinath and Rigobon, 2008), or consumer prices (e.g., Campa and Goldberg, 2010). Consistent with other recent firm-level studies that use exports data for the whole manufacturing sector (e.g., Amiti et al., 2014; Berman et al., 2012; Chatterjee et al., 2013), we find that exchange rate pass-through for export prices is incomplete, but large. Within this literature, most closely related to our work are Antoniades and Zaniboni (2015), Auer and Chaney (2009), Auer et al. (2014), and Bernini and Tomasi (2015) who explore empirically whether pass-through is lower for higher quality goods (only Auer and Chaney, 2009, do not find any evidence to support this relationship). We differ from these papers along several dimensions. First, we document the pricing-to-market behavior of multi-product exporters that leads to incomplete passthrough. Instead, Antoniades and Zaniboni (2015), Auer and Chaney (2009), and Auer et al. (2014) do not use firm-level data, and only study pass-through by using import and retail prices which contain nontradables. Second, our analysis relies on an observable measure of quality, while Auer et al. (2014) estimate hedonic quality indices, Antoniades and Zaniboni (2015) and Auer and Chaney (2009) infer product quality from data on prices and unit values, and Bernini and Tomasi (2015) replicate the Khandelwal (2010) procedure. Third, in contrast to these papers, our focus on wine allows us to control for the potential endogeneity of quality. Finally, we simultaneously analyze the response of unit values and quantities differentiated by quality to exchange rate movements, while these papers only study price behavior. 4

ACCEPTED MANUSCRIPT

RI P

T

Second, as pricing-to-market and incomplete pass-through lead to deviations from the Law of One Price, our paper relates to studies that examine the behavior of product-level real exchange rates. Recent contributions that rely on highly disaggregated product-level data include Simonovska (2015) who studies the positive relationship between the price of tradable goods and per capita income. Cavallo, Neiman, and Rigobon (2014) find that the Law of One Price holds very well within, but not outside, of currency unions. Burstein and Jaimovich (2012) show that product-level real exchange rates across countries are four times as volatile as nominal exchange rates.

PT

ED

MA

NU

SC

Third, this paper contributes to the growing literature on quality and trade, which until recently mostly relied on trade unit values to measure quality.6 At the country-level, Hummels and Klenow (2005) and Schott (2004) show that export unit values increase with exporter per capita income, while Hallak (2006) and Hummels and Skiba (2004) find that richer countries have a stronger demand for high unit value exporting countries. Feenstra and Romalis (2014) decompose the unit values of internationally traded goods into quality and quality-adjusted prices, and show that developed countries both export and import more of higher quality goods. Other papers investigate how quality relates to the performance of exporters using firm-level data. Manova and Zhang (2012a) focus on Chinese firm-level export prices, and find evidence of quality sorting in exports. Kugler and Verhoogen (2012), Manova and Zhang (2012b), and Verhoogen (2008) highlight the correlation between the quality of inputs and of outputs focusing on Mexican, Chinese, and Colombian firms, respectively. Bastos, Silva, and Verhoogen (2014) show that exporting to richer countries leads Portuguese firms to upgrade the quality of their outputs, which requires buying higher quality inputs. Amiti and Khandelwal (2013) and Fan, Li, and Yeaple (2014) show that lower import tariffs increase export quality.7

AC

CE

The paper is organized as follows. Section 2 presents the theoretical framework, and discusses how the features of the wine industry can be reconciled with the main assumptions of the model. Section 3 describes the firm-level exports customs data, the wine experts quality ratings, and the macroeconomic data we use. Section 4 presents our main empirical results. Section 5 addresses endogeneity. Section 6 discusses extensions while Section 7 provides robustness checks. Section 8 concludes.

2

Theoretical Framework

Heterogeneous pricing-to-market due to differences in product quality arises in several models based on endogenous and variable markups. Overall, by showing that pass-through decreases with quality, our empirical analysis validates all these classes of models.8 Atkeson and Burstein (2008) develop a model where firms in each sector produce and export a differentiated good. Consumers in each market have a nested CES demand over the varieties of goods, and the elasticity of substitution across the varieties within sectors is larger than the one across sectors. In this setting, higher quality represents a higher export market share for a firm at a given price, and 6

This approach is criticized by, among others, Khandelwal (2010) who compares exporters’ market shares conditional on price to infer the quality of exports. 7 Also, see Baldwin and Harrigan (2011), Bastos and Silva (2010), Hallak and Schott (2011), Hallak and Sivadasan (2013), and Johnson (2012). 8 In contrast, using a translog expenditure function to generate endogenous markups in a model where firms are heterogeneous in productivity and product quality, Rodr´ıguez-L´ opez (2011) predicts that the response of markups to exchange rate changes falls with productivity and quality, therefore pass-through increases with productivity and quality.

5

ACCEPTED MANUSCRIPT

T

due to imperfect competition ` a la Cournot, high market share firms perceive a lower demand elasticity and have higher markups. As a real depreciation expands market shares, high quality and therefore high market share firms price-to-market more than others. Importantly, the model predicts that quality affects pass-through only through market shares (see, also, Amiti et al., 2014).

SC

RI P

The second model is Melitz and Ottaviano (2008) with heterogeneous firms, fixed costs of entry, and non-CES preferences. Under the assumption that high performance firms are also high quality, this setup delivers a demand elasticity with respect to the final consumer price that decreases with quality. In response to a real depreciation, the prices faced by consumers fall and all exporters react by increasing their markups so there is pricing-to-market and incomplete pass-through. However, higher quality firms increase their markups more than others.9

ED

MA

NU

In Auer et al. (2014), firms in vertically differentiated industries compete monopolistically in the quality space. They sell goods of heterogeneous quality to consumers with non-homothetic preferences who differ in their income, and thus in their willingness to pay for a marginal increase in quality. High market share firms perceive a lower demand elasticity and have higher markups. As a depreciation increases market shares, high market share firms charge higher markups. High quality firms have relatively higher market shares in higher income countries, therefore the model predicts that passthrough decreases with quality in richer countries, but rises with quality in poorer countries.

2.1

AC

CE

PT

Finally, the model of Corsetti and Dedola (2005) with additive (per unit) distribution costs is the one we extend in Section 2.1 by allowing multi-product firms to export products with heterogeneous levels of quality.10 In this setting, a real depreciation decreases the import price in the currency of the destination country, but as distribution costs, paid in local currency, are unaffected by changes in exchange rates, the share of the consumer price that depends on the export price falls. This reduces the demand elasticity perceived by exporters to their export price, which allows all firms to increase their markups. However, as higher quality goods face higher distribution costs, pricing-to-market is predicted to be stronger for higher quality products.

A Model of Pricing-to-Market and Quality

The Home country (Argentina in our case) exports to multiple destinations in one sector characterized by monopolistic competition. The representative agent in destination country j has preferences over the consumption of a continuum of differentiated varieties given by11 U (Cj ) =

Z

[s(ϕ)xj (ϕ)] Ψ

σ−1 σ

dϕ



σ σ−1

,

(1)

9 Based on Melitz and Ottaviano (2008), Basile, de Nardis, and Girardi (2012) predict that pass-through decreases with quality, while Yu (2013) shows that incomplete pass-through results from firms adjusting both their markups and the quality of their products in response to changes in exchange rates. 10 By allowing firms to export multiple products with heterogeneous levels of quality, where higher quality is associated with higher marginal costs, our model differs from Berman, Martin, and Mayer (2012) where firms are heterogeneous in productivity or quality but only export one good. It also differs from Chatterjee, Dix-Carneiro, and Vichyanond (2013) who model firms as multi-product but rank the goods exported by each firm in terms of efficiency rather than quality, where higher efficiency implies lower marginal costs. For other studies that rely on the Corsetti and Dedola (2005) framework to generate heterogeneous pass-through, see Antoniades and Zaniboni (2015) and Bernini and Tomasi (2015). 11 For similar preferences, see Baldwin and Harrigan (2011), Berman et al. (2012), Crozet, Head, and Mayer (2012), Fan, Li, and Yeaple (2014), Johnson (2012), Kugler and Verhoogen (2012), and Manova and Zhang (2012b).

6

ACCEPTED MANUSCRIPT

T

where xj (ϕ) is the consumption of variety ϕ, s(ϕ) the quality of variety ϕ, and σ > 1 the elasticity of substitution between varieties.12 The set of available varieties is Ψ. Quality captures any intrinsic characteristic or taste preference that makes a variety more appealing for a consumer given its price. Therefore, consumers love variety, but also quality.

NU

SC

RI P

Firms are multi-product and heterogeneous in product quality. The parameter ϕ, which denotes each variety, indicates how efficient each firm is at producing each variety, therefore ϕ has both a firm- and a product-specific component. Each firm produces one “core” product, but in contrast to Chatterjee et al. (2013) or Mayer, Melitz, and Ottaviano (2014) who consider that a firm’s core competency lies in the product it is most efficient at producing – and has lower marginal costs – we assume that a firm’s core competency is in its product of superior quality which entails higher marginal costs.13

ED

MA

The efficiency associated with the core product is given by a random draw Φ, therefore each firm is indexed by Φ. Let us denote by r the rank of the products in increasing order of distance from the firm’s core, with r = 0 referring to the core product with the highest quality. Firms therefore observe a hierarchy of products based on their quality levels. A firm with core efficiency Φ produces a product r with an efficiency level ϕ given by ϕ (Φ, r) = Φϑr , (2)

PT

where ϑ > 1. Products with smaller r (higher quality) are closer to the core, and have a lower efficiency ϕ (Φ, r). As in Berman et al. (2012), higher quality goods have a lower efficiency because they have higher marginal costs in the form14

CE

s(ϕ (Φ, r)) =



w ϕ (Φ, r)

λ

,

(3)

AC

where λ > 1 implies that markups increase with quality, and w is the wage of the Home country. The closer a product is from the core with the highest quality (i.e., the smaller r), the lower is efficiency ϕ (Φ, r), and the higher are marginal costs and quality s(ϕ (Φ, r)). Firms face three types of transaction costs: an iceberg trade cost τj > 1 (between Home and destination j), a fixed cost of exporting Fj (which is the same for all firms and products and only depends on destination j), and an additive (per unit) distribution cost in destination j.15 The latter captures any wholesale and retail costs associated with the selling, storage, marketing, advertising, or transport of goods paid in the currency of the destination country. If distribution requires ηj units of labor in country j per unit sold, and wj is the wage rate in country j, distribution costs are given by ηj wj s (ϕ(Φ, r)). As in Berman et al. (2012), we assume that higher quality goods have higher (per unit) distribution costs. Most importantly, as distribution is outsourced so that distribution costs are paid in the currency of the importing country, they are unaffected by changes in the exchange rate.16 12

As in Kugler and Verhoogen (2012), quality is treated as single-dimensional and is therefore only chosen by firms. Higher quality goods are assumed to have higher marginal costs because they require higher quality, and therefore more expensive, inputs. See Crin` o and Epifani (2012), Fan et al. (2014), Feenstra and Romalis (2014), Hallak and Sivadasan (2013), Johnson (2012), Kugler and Verhoogen (2012), Manova and Zhang (2012a), and Verhoogen (2008). 14 See Crin` o and Epifani (2012) and Hallak and Sivadasan (2013) for models where marginal costs decrease with firm-level productivity and increase with product quality. 15 We do not model entry and exit in export markets. See Campos (2010) for the effects of entry and exit on pass-through. 16 The predictions of the model continue to hold if distribution costs are not paid by the exporting firm. 13

7

ACCEPTED MANUSCRIPT In units of currency of country j, the consumer price in j of a variety exported from Home to j is pcj (ϕ) ≡

pj (ϕ(Φ, r))τj + ηj wj s (ϕ(Φ, r)) , εj

(4)



pj (ϕ(Φ, r)τj + ηj w j s (ϕ(Φ, r)) εj

−σ

,

(5)

NU

xj (ϕ) =

Yj Pjσ−1

SC

RI P

T

where pj (ϕ(Φ, r)) is the export price of the good exported to j, expressed in Home currency, and εj is the nominal exchange rate between Home and j. It is straightforward to see that any change in the exchange rate εj leads to a less than proportional change in the consumer price pcj (ϕ) (i.e., incomplete pass-through) given that local distribution costs are unaffected by currency fluctuations. The quantity demanded for this variety in country j is

MA

where Yj and Pj are country j’s income and aggregate price index, respectively.17 The costs, in the currency of the Home country, of producing xj (ϕ) τj units of each good (inclusive of transportation costs) and of selling them to country j are cj (ϕ) =

wxj (ϕ(Φ, r)) τj + Fj . ϕ(Φ, r)

(6)

PT

ED

Expressed in Home currency, the profit maximizing export price for each product the firm exports to country j is   ηj qj ϕ(Φ, r)s (ϕ(Φ, r)) σ w w pj (ϕ) = 1+ = m (ϕ(Φ, r)) , (7) σ−1 στj ϕ(Φ, r) ϕ(Φ, r)

AC

CE

where qj ≡ εj wj /w is the real exchange rate between Home and j. In contrast to the standard DixitStiglitz markup (Dixit and Stiglitz, 1977), the presence of local distribution costs leads to variable σ , increase with quality s (ϕ(Φ, r)), markups m (ϕ(Φ, r)) over marginal costs that are larger than σ−1 the real exchange rate qj (i.e., a real depreciation), and local distribution costs ηj . The volume of exports xj (ϕ) is given by xj (ϕ) =



σ−1 σ

σ

Yj Pjσ−1



w τj + η j w j ϕ(Φ, r)s(ϕ(Φ, r))εj

−σ

(8)

therefore the elasticity (in absolute value) of the exporter’s demand xj (ϕ) with respect to the export price pj (ϕ) is ∂xj (ϕ) pj (ϕ) στj + ηj qj ϕ(Φ, r)s(ϕ(Φ, r)) = ej = , (9) ∂pj (ϕ) xj (ϕ) τj + ηj qj ϕ(Φ, r)s(ϕ(Φ, r))

which decreases with quality and a real depreciation.18 For a product that is closer to the core, quality is higher, the elasticity of demand is lower, and the markup is higher. The model leads to two testable predictions on the effects of exchange rate changes on export prices and quantities.  1 R The aggregate price index is given by Pj = Ψ pj (ϕ(Φ, r))1−σ dϕ 1−σ 18 Instead, and in contrast to the Melitz and Ottaviano (2008) model, the demand elasticity with respect to the consumer price does not vary with quality. 17

8

ACCEPTED MANUSCRIPT Prediction 1 The firm- and product-specific elasticity of the export price pj (ϕ) to a change in the real exchange rate qj , denoted by epj , increases with the quality of the exported good, s(ϕ(Φ, r)):

T

epj

∂pj (ϕ) qj ηj qj ϕ(Φ, r)s(ϕ(Φ, r)) = . = ∂qj pj (ϕ) στj + ηj qj ϕ(Φ, r)s(ϕ(Φ, r))

∂xj (ϕ) qj στj = = . ∂qj xj (ϕ) τj + ηj qj ϕ(Φ, r)s(ϕ(Φ, r))

SC

e xj

RI P

Prediction 2 The firm- and product-specific elasticity of the volume of exports xj (ϕ) to a change in the real exchange rate qj , denoted by exj , decreases with the quality of the exported good, s(ϕ(Φ, r)):

2.2

PT

ED

MA

NU

Intuitively, the mechanism is the following. A real depreciation decreases the import price in the currency of the destination country, and as distribution costs are unaffected by changes in exchange rates, the share of the consumer price that depends on the export price falls. This reduces the elasticity of demand perceived by exporters to their export price, which allows all firms to increase their markups. As higher quality goods face higher distribution costs, a large share of their final consumer price does not depend on the export price. Therefore, the perceived elasticity of demand to the export price decreases with quality. As a result, the markups of higher quality goods can be increased by more than for lower quality goods in response to a real depreciation. This leads to heterogeneous pricingto-market which is stronger for higher quality goods. In turn, this implies that the response of export volumes to a real depreciation decreases with quality.

Cross-Country Heterogeneity in the Preference for Quality

AC

CE

Evidence suggests that consumer preferences for quality vary from one country to the other as preferences are affected by per capita income. In particular, consumers in richer countries are assumed to have a stronger preference for higher quality goods.19 We therefore extend the model of the previous section and allow for non-homothetic preferences in quality. The utility function becomes U (Cj ) =

Z

[s(ϕ, yj )xj (ϕ)] Ψ

σ−1 σ

dϕ



σ σ−1

,

(10)

where we replace s(ϕ) by s(ϕ, yj ) to capture that the preference for quality can increase in per capita income, yj . Local distribution costs ηj wj s(ϕ, yj ) are thus higher in higher income countries.20 This allows us to derive two additional testable predictions. Prediction 3 The firm- and product-specific elasticity of the export price pj (ϕ) to a change in the real exchange rate qj , denoted by epj , increases with the quality of the exported good s(ϕ(Φ, r), yj ), and by more for higher income than for lower income destination countries: ∂pj (ϕ) qj ηj qj ϕ (Φ, r) s(ϕ(Φ, r), yj ) = . epj = ∂qj pj (ϕ) στj + ηj qj ϕ (Φ, r) s(ϕ(Φ, r), yj ) 19

Hallak (2006) finds that rich countries import more from countries that produce higher quality goods. In Fajgelbaum, Grossman, and Helpman (2011), the share of consumers who buy higher quality goods rises with income. Also, see Crin` o and Epifani (2012), Feenstra and Romalis (2014), Manova and Zhang (2012a), Simonovska (2015), and Verhoogen (2008). 20 Dornbusch (1989) shows that the prices of services are lower in poorer than in richer countries, suggesting that local distribution costs are lower in poorer countries.

9

ACCEPTED MANUSCRIPT

RI P

T

Prediction 4 The firm- and product-specific elasticity of the volume of exports xj (ϕ) to a change in the real exchange rate qj , denoted by exj , decreases with the quality of the exported good s(ϕ(Φ, r), yj ), and by more for higher income than for lower income destination countries: ∂xj (ϕ) qj στj = exj = . ∂qj xj (ϕ) τj + ηj qj ϕ (Φ, r) s(ϕ(Φ, r), yj )

Relevance for Wine

MA

2.3

NU

SC

The elasticity of the exporter’s demand with respect to the export price is now decreasing in quality, per capita income, and with a real depreciation. In response to a real depreciation, firms can increase their markups by more for higher quality goods, but also when exporting to higher income countries. In turn, this implies that the response of volumes to exchange rate changes decreases with quality and per capita income. The assumption that consumers in richer countries are less price-elastic than consumers in poorer countries – which enables firms to charge higher markups in richer destinations – is consistent with Alessandria and Kaboski (2011) and Simonovska (2015).

ED

The predictions of our model, which features monopolistic competition and local distribution costs, depend crucially on the assumptions that higher quality wines have higher (per unit) distribution costs, higher marginal costs, and higher markups. This section discusses how the characteristics of the wine industry can be reconciled with these assumptions.21

CE

PT

In terms of market structure, the wine industry is composed of many producers offering a large number of highly differentiated products for which consumers are willing to pay a wide range of different prices. Because of product differentiation, wine producers have market power. Therefore, the sector can be assumed to operate in a monopolistically competitive environment (Thornton, 2013).

AC

To generate incomplete pass-through, our model features the role of local distribution costs which can be interpreted as any (per unit) cost related to the selling, marketing, advertising, or transport of goods paid in the destination’s currency. In the case of wine, distribution costs are important. For instance, in the US, about 90 percent of wine is distributed through a three-tier distribution system, whereby wine is sold by a producer or by an importer to a distributor, who then sells it to a retailer, who finally sells it to consumers. Each tier involves a margin to compensate for each activity that is carried out. The margin of the distributor is around 30 percent, and covers the costs related to transport, storage, and promotional activities. In turn, the retailer margin varies between 30 percent for grocery stores and wine shops and 200 percent for restaurants, bars, and hotels (Thornton, 2013). The prediction that pass-through decreases with quality depends crucially on the assumption that higher quality goods have higher (per unit) distribution costs. In the case of wine, this assumption 21

As our assumptions are likely to hold more generally, the predictions of our model extend beyond the case of wine. First, local distribution costs are large. They represent between 40 and 60 percent of cross-country retail prices (Burstein, Neves, and Rebelo, 2003), and between 30 and 50 percent of the total costs of goods exported by OECD countries (Campa and Goldberg, 2010). Second, (per unit) distribution costs are likely to rise with quality. For instance, higher quality goods require heavier packaging, more careful and costly handling, face higher insurance fees (Hummels, 2007; Hummels and Skiba, 2004), and higher advertising and marketing costs (e.g., Kugler and Verhoogen, 2012; Phillips, Chang, and Buzzell, 1983; Sutton, 1998). Finally, evidence shows that higher quality goods have higher marginal costs as they require higher quality, and therefore more expensive, inputs (e.g., Kugler and Verhoogen, 2012; Manova and Zhang, 2012b; Verhoogen, 2008), and have higher markups (e.g., Auer et al., 2014; Bellone, Musso, Nesta, and Warzynski, 2016; De Loecker and Warzynski, 2012; Feenstra and Levinsohn, 1995).

10

ACCEPTED MANUSCRIPT

SC

RI P

T

is plausible. First, higher quality wines are more costly to transport as they are typically shipped in temperature-controlled containers, they need heavier bottles and cartons for shipping, require more careful and therefore more costly handling, and face higher insurance fees (Hummels, 2007; Hummels and Skiba, 2004). Second, tasting events to promote wines are organized by distributors for higher quality wines only, which therefore have higher marketing and advertising costs. Third, top quality wines need to be stored in refrigeration units to maintain their temperature, and are therefore subject to higher storage costs. Finally, top quality wines are more likely to be sold in specialized wine stores than in supermarkets, and the former usually charge a higher price because of the professional customer service they offer. In other words, retail costs are higher for higher quality wines.

AC

CE

PT

ED

MA

NU

Higher quality wines can also be expected to entail higher marginal costs (Thornton, 2013). First, the quality of wine depends predominantly on the quality of the grapes. Although grape quality is mostly affected by geography and weather-related factors, achieving higher quality grapes can be more costly as it requires, for instance, to restrict the grape yield per acre of vineyard, to grow the grapes organically, to use labor-intensive treillis systems, and to prune, trim, pick, and sort the grapes more carefully which in turn necessitates more skilled, manual labor. Higher quality wines also require higher quality, and therefore more costly inputs such as the additives and yeast types used during the various stages of fermentation or as preservatives. Second, achieving higher quality wines depends on the methods of production chosen, including the way the wine is extracted from the grapes (the manual labor-intensive “punching down” technique is more costly but results in superior wine quality), or the type of vessel to be used for ageing and fermentation. Compared to stainless-steel tanks, oak barrels increase quality by adding flavor to the wine, but are more expensive due to the cost of the oak (the barrels are between 75 and 90 percent more costly than similar sized stainless-steel tanks, and add between two to three US dollars to the cost of a bottle of wine), their short lifetime (the oak flavors last for three or four vintages only), and the evaporation of the wine due to the porous nature of the oak (around ten percent per year). Quality can be further enhanced by using smaller oak barrels to increase the amount of wood per liter of wine, but these are, in turn, more expensive. Finally, higher quality wines need to mature for a longer period of time, which increases storage costs. As an illustration, Table 1 breaks down into several components the prices of non-EU wines sold in UK retail outlets (Joseph, 2012). We believe that these figures should provide us with some useful insights on the price composition of Argentinean wines sold in the UK. The table shows that more expensive, and hence higher quality wines, have higher retail and distribution costs.22 Also, the amount that goes to the winemaker, which includes his profit and covers the costs of producing the wine, clearly increases with the price, and therefore with quality. Unfortunately, the table does not provide any information on the winemaker markup which is the one that is modeled in the theory. However, anecdotal evidence suggests that the producer markup is also likely to increase with the price/quality of the wine: for a five pound sterling wine sold on the UK market, the producer markup is estimated to be around 40 pence, and increases to about ten pounds for a 25 pound bottle.23 22 23

Shipping costs are stable at 13 pence per bottle but may rise with the heavier bottles used for more expensive wines. See www.thirtyfifty.co.uk.

11

ACCEPTED MANUSCRIPT 3

Data and Descriptive Statistics

Firm-Level Customs Data

RI P

3.1

T

Our data set gathers information from different sources: firm-level customs data, wine experts quality ratings, and macroeconomic data.

NU

SC

Firm-level exports are from the Argentinean customs and are provided to us by a private vendor called Nosis. For each export flow we observe the name of the exporting firm, the country of destination, the date of declaration, the 12-digit HS classification code, the FOB value of exports (in US dollars), and the volume (in liters) exported between 2002 and 2009.24 We also have the name of the wine exported, its type (red, white, or ros´e), grape (Malbec, Chardonnay, etc.), and vintage year. Figure 1 compares the total value of Argentina’s wine exports from the customs data set with the value reported in the Commodity Trade Statistics Database (Comtrade) of the United Nations (HS code 22.04). The data coincide extremely well.

ED

MA

Given that actual export prices are not available, we proxy for them using the unit values of exports in the currency of the exporter, computed as the ratio between the export value in Argentinean pesos and the corresponding export volume in liters. To convert the value of exports (in US dollars) into pesos, we use the peso to US dollar exchange rate in the month in which the shipment took place. We then aggregate the data at an annual frequency.25

AC

CE

PT

We clean up the data in several ways. First, we drop any wine for which either the name, grape, type, or vintage year is missing, cannot be recognized, or is classified as “Undefined.” As sparkling wines, dessert wines, and other special varieties do not have any vintage year, they are excluded from the data set. Second, we only keep the export flows recorded as FOB, which therefore exclude shipping costs, tariffs, or distribution and retail costs charged in the destination country. Third, as we are interested in how quality affects the pricing and export decisions of wine producers, we restrict our analysis to the manufacturing sector, and drop wholesalers and retailers. Fourth, due to the rise in inflation in 2007 and increasing worries of capital outflows, the Argentinean government introduced some restrictions to obtain import licenses which culminated in an extreme measure whereby firms were authorized to import only if they exported goods for an equivalent value to those they intended to import.26 We therefore drop from the sample the firms unrelated to the wine industry that exported 24

Due to confidentiality reasons, the customs data cannot make the name of the exporter public, therefore Nosis uses its own market knowledge and an algorithm to identify a first, a second, and a third probable exporter. To determine the exporter’s identity, for each wine name we gathered from the Instituto Nacional de Vitivinicultura (INV) the name of its producer and of the wholesaler/retailer authorized to export this wine. For each wine name, we then compared the names of the probable exporters reported by Nosis with the names reported by the INV. We could identify 86, five, and one percent of the first, second, and third probable exporters, respectively, while the remaining eight percent could not be matched. In our sample, these proportions become 96.5, three, one half, and zero percent, respectively. 25 Our results still hold when measuring unit values, export volumes, and exchange rates at a quarterly frequency. The reason why we aggregate the data at an annual frequency is that many of the other variables we use, such as the real effective exchange rates or real GDPs per capita, are not available at a quarterly frequency. 26 See http://www.ustr.gov/sites/default/files/2013%20NTE%20Argentina%20Final.pdf. The World Trade Organization has subsequently ruled that the requirement that imports should be balanced with exports violates global trade rules (The Financial Times, “WTO Rules Against Argentine Import Curbs,” 22 August 2014).

12

ACCEPTED MANUSCRIPT

RI P

T

wine over the period.27 Fifth, we drop the wines that are not produced in Argentina.28 Sixth, we drop a number of typos which we are unable to fix. We exclude a few cases where the vintage year reported is ahead of the year in which the export took place, and the observations where the value of exports is positive, but the volume is zero. Finally, we exclude some outliers: for each exporter, we drop the observations where unit values are larger or smaller than 100 times the median unit value charged by the firm.

3.2

PT

ED

MA

NU

SC

Recent papers on heterogeneous pass-through typically define a “product” according to trade classifications such as the CN or the HS (e.g., Amiti et al., 2014; Auer and Chaney, 2009; Berman et al., 2012; Chatterjee et al., 2013). As Table 2 shows, the 6-digit HS classifies wines into four different categories, depending on whether the wines are sparkling or not, and on the size of the containers in which they are shipped (i.e., larger or smaller than two liters). Argentina further disaggregates the HS classification at the 12-digit level, but this only enlarges the number of different categories, or “products,” to eleven. The problem is that changes in unit values defined at this level may reflect compositional or quality changes rather than price changes as there may be more than one distinct product within a single HS code (Feenstra and Romalis, 2014). In contrast, the detail provided by our data allows us to define a product as a combination between a wine name, type, grape, vintage year, and the capacity of the container used for shipping (identified using the HS code), so that compositional or quality changes are unlikely to affect unit values.29 Our cleaned sample includes a total of 21,647 different wines of which 6,720 can be matched with quality ratings. The 6,720 wines represent 43 percent of the total FOB value of red, white, and ros´e wine exported between 2002 and 2009.

Quality

AC

CE

The editors of the Wine Spectator magazine review more than 15,000 wines each year in blind tastings, and publish their ratings in several issues throughout the year. The ratings are given on a (50,100) scale according to the name of the wine, its grape, type, and vintage year, which are characteristics we all observe in the customs data set. A larger score implies a higher quality.30 Table 3 lists the six different categories the wines fall in depending on their scores. We match the wines from the customs data set with the ones reviewed by the Wine Spectator by name, type, grape, and vintage year, and each wine is assigned a single quality rating.31 We end up 27

We discard a total of 157 firms including wine wholesalers (20), retailers (2), unidentified firms (70), as well as firms from the sectors of agriculture (5), soft drinks or malt liquors (6), textiles (6), chemicals (6), metals (1), paints, glass, and paper (3), machinery (1), motorcycles (1), and services (25) such as consultancy or book publishing firms. These firms represent 12,376 observations in the raw data set (at the transaction-level), or 1.6 percent of the total value of wine exports between 2002 and 2009. Of the 157 firms, only 41 export wines that can be matched with the Wine Spectator ratings, and these firms represent 0.17 percent of the total value of exports over the period (or 1,520 observations). In the data set aggregated at a yearly frequency, the 41 firms represent 345 observations. 28 These wines represent 0.7 percent of the total FOB value of wine exported between 2002 and 2009. 29 In our sample, each wine is produced, and exported, by one firm only. In contrast, a product defined at the CN- or HS-levels can be produced, and therefore exported, by more than one firm. 30 Quality depends on the amounts of characteristics embodied in a wine, and is assessed by evaluating those characteristics. Subjective quality refers to the evaluation of consumers based on personal tastes and preferences, whereas objective quality assesses the characteristics that are independent of personal preferences such as balance, complexity, finish, concentration, or flavor intensity. The aim of wine critics is to provide an objective assessment of wine quality, and they may therefore conclude that a wine is of high quality even though they personally dislike it (Thornton, 2013). However, as we discuss later, some argue that experts ratings can, to some extent, be subjective. 31 The quality scores are time-invariant. Although evidence suggests that the quality of Argentinean wines (and their

13

ACCEPTED MANUSCRIPT

T

with 6,720 wines exported by 209 firms over the 2002–2009 period. Our approach to measuring quality is similar to Crozet et al. (2012) for Champagne. However, these authors are unable to distinguish between the different varieties sold by each firm, and therefore assume that each firm exports one type of Champagne only. In addition, their ratings are only measured on a (1,5) scale, with a larger value indicating a higher quality.

MA

NU

SC

RI P

We rely on the Wine Spectator for our baseline regressions because it has the largest coverage of Argentinean wines. However, we check the sensitivity of our results using an alternative rating produced by Robert Parker. Parker is a leading US wine critic who assesses wines based on blind tastings, and publishes his ratings in a newsletter, the Wine Advocate.32 His rating system also employs a (50,100) point scale, with wines being ranked according to their name, type, grape, and vintage year, and with a larger value indicating a higher quality.33 Table 3 lists the different categories considered by Parker. Compared to the Wine Spectator, the scores are slightly more generous (for instance, a wine ranked 74 is “Not recommended” by the Wine Spectator, but is “Average” according to Parker). The customs data and the Parker ratings are matched for 3,969 wines exported by 181 firms.

AC

CE

PT

ED

Figure 2 plots the Wine Spectator against the Parker ratings for the 2,433 wines (exported by 135 firms) that have scores from both sources. For the sake of clarity, for both ratings quality is measured between one and six, where each value corresponds to one of the different bins defined in Table 3. A value of one indicates that the wine is either “Not recommended” or “Unacceptable,” while a value of six that the wine is “Great” or “Extraordinary.” For Parker, notice that only the three highest quality bins are depicted in the graph.34 Each circle is sized proportionately to the number of wines at each point in the figure. Most wines are rated as “Above average/very good” (a value of four) by Parker, and are either “Good” or “Very good” (a value of three or four) according to the Wine Spectator, again suggesting that Parker is, on average, more generous in its ratings. A relatively large number of wines are rated as “Outstanding” (a value of five) by both tasters, and a few as “Great” or “Extraordinary” (a value of six). Notice that a few wines are ranked as “Not recommended” (a value of one) by the Wine Spectator, but as “Outstanding” (a value of five) by Parker. The two ratings are, however, positively correlated. The Pearson correlation is equal to 0.51, while Kendall’s correlation index of concordance between the two ratings is 0.47.

3.3

Macroeconomic Data

The data on real GDPs per capita (in US dollars) are from the Penn World Tables, and the consumer price indices (CPI) and nominal exchange rates are from the International Financial Statistics (IFS) of the International Monetary Fund (IMF). The real exchange rate is defined as the ratio of consumer price indices times the average yearly nominal exchange rate, and an increase indicates a real depreciation of the Argentinean peso. The real effective exchange rates are from the IFS and the Bank of International price) is not much affected by ageing (Thornton, 2013), in some regressions we include vintage year fixed effects to control for product characteristics. When explaining unit values, the coefficients on the vintage year fixed effects are larger for older wines, suggesting that older wines tend to be more expensive. A higher price for older wines could indicate that quality improves with ageing, but may also reflect the costs of storing the wines over time. 32 Both the Wine Spectator and Parker are US-based ratings. Other ratings have a poor coverage of Argentinean wines. 33 Starting with a score of 50, Parker assigns points for appearance (5 points), smell (15 points), taste (20 points), and overall quality and ageing potential (10 points). The Wine Spectator does not make public how the points are awarded. 34 When using Parker in our regressions, four out of the six bins listed in Table 3 are however included in the sample.

14

ACCEPTED MANUSCRIPT Settlements, and an increase indicates a real depreciation. Bilateral distances are from the Centre d’Etudes Prospectives et d’Informations Internationales (CEPII).

3.4

SC

RI P

T

Between 2002 and 2009, Argentina witnessed major nominal and real exchange rate fluctuations. Figure 3 plots the monthly nominal and real exchange rates between the Argentinean peso and the US dollar. After the crisis of 2001, the fixed exchange rate was abandoned and as a result, in 2002, the peso depreciated both in nominal and real terms by up to 75 percent. The export boom that followed lead to a massive inflow of US dollars into the economy, which helped to appreciate the peso against the dollar. The peso then remained stable – but slightly appreciated in real terms – until 2008 when it depreciated again with the advent of the global financial crisis and the rise in domestic inflation.

Descriptive Statistics

CE

PT

ED

MA

NU

Table 4 summarizes our trade data by year, and shows that the exports included in our sample increased threefold between 2002 and 2009. A total of 794 wines were exported by 59 different firms in 2002, while in 2009 this increased to 151 firms exporting 1,833 different wines. Over the whole period, our sample includes 6,720 wines exported by 209 wine producers.35 As shown in Table 5, these firms exported an average of 139 different wines, from a minimum of one to a maximum of 510 (in the sample, 15 firms appear as having exported one wine only; in reality, they exported more than one wine, but only one could be matched with the quality ratings). Exporters charged between two cents and 381 US dollars per liter of wine exported, with an average of five US dollars per liter.36 They exported to an average of 24 different destinations, from a minimum of one to a maximum of 65. The mean Wine Spectator rating is 85, the lowest rated wine receives a score of 55, and the highest a score of 97. The distribution across wines is symmetric as the mean and the median are equal. The Parker scores vary between 72 and 98, with an average of 87 (only four out of the six bins listed in Table 3 are therefore included in our sample). Table 6 shows that, with the exception of Brazil, Argentinean wine producers mostly exported to developed economies, the United States being the top destination market.

AC

Table 7 provides a snapshot of our data. For confidentiality reasons, we can neither report the exporter nor the wine names, therefore these are replaced by numbers and letters instead. The table shows that, whether we use the Wine Spectator or Parker ratings, firms exported wines with heterogeneous levels of quality (between 73 and 92 for Firm 1, and between 85 and 98 for Firm 2). Higher quality wines are sold at a higher price. Besides, the table illustrates that the Law of One Price fails: in 2006, Firm 1 exported the same wine to two different destinations, but charged 13.38 US dollars per liter to the United States, versus 11.09 US dollars per liter to China. Similarly, in 2008, Firm 2 charged 43.27 US dollars per liter to the United Kingdom, versus 24.65 US dollars per liter to Thailand. The differences in the prices of these wines reflect differences in markups as identical wines have the same marginal cost. Interestingly, the prices, and therefore the markups, are higher for exports to higher income countries, which is in line with Simonovska (2015) who argues that firms extract higher markups for identical products when selling to richer countries with lower demand elasticities. Table 8 reports descriptive statistics by quality bin of the Wine Spectator. Wines that are either “Good” or “Very good” represent the largest share of the sample (in terms of number of observations, 35

We observe 882 different wine names, 23 grapes, three types, and 22 vintage years (between 1977 and 2009). In the Trade Unit Values Database (Berthou and Emlinger, 2011), between 2002 and 2009 the 6-digit HS-level FOB unit values of Argentinean wine exports vary between seven cents and 16 US dollars per liter. Given the much higher level of disaggregation of our data, unit values between two cents and 381 US dollars per liter sound reasonable. 36

15

ACCEPTED MANUSCRIPT

T

number of exporters, and the share of exports in the sample). At the other two extremes, “Not recommended” wines, followed by “Great” wines, have the smallest coverages. “Great” wines are more expensive, and are exported to a smaller number of destinations (per wine exported) which are, on average, richer (Hallak, 2006; Verhoogen, 2008), and more distant (Baldwin and Harrigan, 2011; Bastos and Silva, 2010; Feenstra and Romalis, 2014; Hummels and Skiba, 2004; Johnson, 2012).

Empirical Analysis

MA

4

NU

SC

RI P

Finally, Table 9 describes the data across firms. Each firm is allocated to one of the Wine Spectator quality bins based on the highest quality wine the firm exported over the period. Firms exporting higher quality wines tend to export a higher quality, on average.37 Interestingly, these firms also tend to export more variety. Taken together, the 68 firms which exported “Great” and “Outstanding” wines exported “Not recommended” and “Mediocre” wines as well. They exported a larger number of different wines, to a larger number of destinations (per wine exported), and represent the largest share of exports in the sample. At the other extreme, only one firm exported exclusively “Not recommended” wines, while 12 firms exported “Mediocre” wines as their highest quality.

To check if Prediction 1 holds in the data, we estimate the following reduced-form regression

ED

ln U Vijk,t = β1 ln RERj,t + β2 ln RERj,t × qualityk + Dk,t + Dij + ǫijk,t

(11)

CE

PT

where U Vijk,t is the export unit value of wine k exported by firm i to country j in year t, expressed in pesos per liter, and is our proxy for export prices. RERj,t is the real exchange rate between Argentina and country j in year t (an increase in RERj,t indicates a real depreciation of the peso). The quality of wine k is denoted by qualityk , and the Wine Spectator ratings are used for our benchmark specifications.

AC

The export price in the exporter’s currency is a markup over marginal costs. To identify a pricingto-market behavior, which requires markups to respond to exchange rate changes, the regression needs to control for product-specific marginal costs (Knetter, 1989, 1993).38 As assumed in Section 2, higher quality wines can be expected to have higher marginal costs. We therefore perform within estimations, and control for time-varying product-specific marginal costs (which are, as a matter of fact, also firmspecific) by including product-year fixed effects, Dk,t .39 As product-year fixed effects are collinear with quality, the direct effect of quality drops out from the regression. Firm-destination fixed effects, Dij , are also included. The coefficients to be estimated are β1 and β2 , and ǫijk,t is an error term. Prediction 1 requires β2 , the coefficient on the exchange rate interacted with quality, to be positive.40 As all variables are in levels (rather than first differences), the estimated coefficients can be thought of as capturing the long term response of unit values to changes in each of the explanatory variables. In addition, given the level of disaggregation of the data, changes in real exchange rates are assumed 37

A firm can share its expertise in winemaking across all of its products, raising their average quality (Thornton, 2013). Distinguishing between changes in markups and changes in marginal costs is also important from a policy point of view as changes in costs are efficient, while changes in markups are not (Burstein and Jaimovich, 2012). 39 As each wine is only produced in a single year, marginal costs are time-invariant. Still, we include product-year fixed effects to control for any other time-varying costs that may differ across products, such as storage costs. 40 When the interaction between the exchange rate and quality is excluded from (11), β1 captures the degree of pricingto-market. Exchange rate pass-through for export prices is given by (1 − β1 ) × 100. 38

16

ACCEPTED MANUSCRIPT to be exogenous to the pricing decisions of individual firms. Finally, as quality takes on a single value for each product, robust standard errors are adjusted for clustering at the product-level. To test Prediction 2 we estimate (12)

RI P

T

ln Xijk,t = α1 ln RERj,t + α2 ln RERj,t × qualityk + α3 Zj,t + Dk,t + Dij + ǫijk,t

Baseline Results

NU

4.1

SC

where Xijk,t is the FOB export volume (in liters) of wine k exported by firm i to country j in year t. As in gravity models, we include destination-year-specific variables Zj,t , including the importing country’s real GDP per capita and real effective exchange rate as a proxy for country j’s price index (Berman et al., 2012). Prediction 2 requires the coefficient on the interaction term, α2 , to be negative.

PT

ED

MA

Panel A of Table 10 reports the results for unit values. To get a sense of how quality affects unit values, and therefore to allow the coefficient on quality to be estimated, we first estimate equation (11) replacing the product-year fixed effects by firm-year dummy variables (to control, among other factors, for the time-varying marginal costs, reputation, or import intensity of each exporter), and simultaneously control for product characteristics by including grape, type, vintage year, HS, and province of origin of the grapes fixed effects.41 Fixed effects for the wine names are not included as they are collinear with the firm fixed effects (as each wine is exported by one producer only). Column (1) only includes the exchange rate and quality as regressors, and shows that higher quality wines are exported at a higher price, which is consistent with equation (7) and with Crozet et al. (2012) for Champagne. Also, as expected, exporters increase their export prices in response to a real depreciation.

AC

CE

Column (2) includes product-year fixed effects, therefore quality drops out from the regression. When the real exchange rate fluctuates, exporters change their export prices: in response to a ten percent real depreciation, they raise their prices (in pesos) by 1.9 percent, therefore pricing-to-market is low and pass-through for export prices is large at 81 percent.42 This large degree of pass-through that we find for the wine industry is consistent with the findings of other papers that use firm-level exports data for the whole manufacturing sector.43 For instance, pass-through for export prices is estimated at 79 percent for Belgian exporters (Amiti et al., 2014), 92 percent for French exporters (Berman et al., 2012), 97 percent for Italian exporters (Bernini and Tomasi, 2015), 77 percent for Brazilian exporters (Chatterjee et al., 2013), and at 96 percent for Chinese exporters (Li, Ma, and Xu, 2015).44 41 “Old World” wines produced in Europe are labelled by the region where they are made and where the grapes are grown. In contrast, “New World” wines are classified by the grape variety used for producing the wine, and the grapes can be grown in different provinces. This is the reason why we control for the province of origin of the grapes. According to Thornton (2013), the location where the grapes are grown may also affect wine prices through its reputation. 42 One possible reason for low pricing-to-market, and high pass-through, is transfer pricing between multinational firms exporting to foreign subsidiaries. If markups are not adjusted by exporters but by their subsidiaries in the destination country, pass-through can be high for export prices, but low for consumer prices (Knetter, 1992). This scenario is unlikely in our case as our sample includes wine producers only and excludes multinational firms. 43 In the Chen and Juvenal (2013) working paper, we aggregated our product-level data at the firm- and at the countrylevels, and estimated exchange rate pass-through at each level of data aggregation. We found that the more aggregated the data, the lower is pass-through, suggesting that aggregate pass-through estimates suffer from an aggregation bias. 44 We checked if exporters adopt different pricing strategies depending on whether the Argentinean peso appreciates or depreciates in real terms, and found that unit values respond significantly more to appreciations than to depreciations. This asymmetric pattern is consistent with firms trying to maintain export market shares by reducing the prices of their exports which become less competitive when the peso appreciates, and with Marston (1990) who shows that pricing-tomarket by Japanese firms is stronger when the Japanese yen appreciates. See, also, Li et al. (2015).

17

ACCEPTED MANUSCRIPT

SC

RI P

T

The estimated coefficient on the exchange rate in column (2) however hides some heterogeneity in the degree of pass-through across products. To see this, column (3) interacts the exchange rate with quality. Its coefficient is positive and significant, providing direct support to Prediction 1. In other words, pricing-to-market rises with quality, while pass-through decreases with quality. The exchange rate elasticity is equal to 0.183 (significant at the one percent level) for a wine at the 5th percentile of the quality distribution (i.e., a “Mediocre” wine with a quality rating of 79), and increases to 0.202 (significant at the one percent level) for a wine at the 95th percentile of the quality distribution (i.e., an “Outstanding” wine with a score of 91), which represents a 10.3 percent increase in pricing-to-market. The effect of quality on pass-through is therefore smaller than in Auer et al. (2014) who find that the pass-through rate for a low quality car is about four times larger than for a top quality car. In column (4), quality is measured using the Parker ratings, and qualitatively, our results continue to hold.45

PT

ED

MA

NU

Panel B reports the results of estimating equation (12) for export volumes. In column (1), where the product-year fixed effects are replaced by firm-year, grape, type, vintage year, HS, and province dummy variables, exports increase with a real depreciation. The elasticity is large and equal to 1.378, which is consistent with evidence that the trade elasticities for emerging economies are larger than for developed countries (Haltmaier, 2011). Once we control for product-year fixed effects in column (2), the exchange rate elasticity becomes insignificant. However, once we allow this elasticity to vary with quality, column (3) shows that, consistent with Prediction 2, the response of export volumes to changes in exchange rates decreases with quality. The exchange rate elasticity of export volumes is equal to 0.858 at the mean value of quality. It is equal to 0.827 for a wine at the 95th percentile of the quality distribution, and increases to 0.888 for a wine at the 5th percentile of the quality distribution. The results remain qualitatively comparable in column (4) where the Parker ratings are used.

AC

CE

In column (1) of Panel B, notice that the coefficient on quality is negative and significant, which contrasts with the positive relationship between trade and quality that is usually estimated in the literature (e.g., Crozet et al., 2012). This negative coefficient is also inconsistent with the expectation that quality “pays,” and that firms should receive higher revenues for higher quality wines (in which case the sum of the elasticities of unit values and export volumes to quality in column 1 should be positive). Crozet et al. (2012) argue that OLS estimations that only include positive trade flows suffer from a selection bias which leads to underestimate the effect of quality on exports. The bias arises because a low quality firm that succeeds to export must be characterized by above-average determinants of export profitability which are unobserved. One way to address sample selection would be to implement Heckman’s correction, as in Helpman, Melitz, and Rubinstein (2008), but this approach would require a variable that explains a firm’s destination-specific fixed costs of exporting, but not its variable trade costs. Missing any such variable, we follow Crozet et al. (2012) and regress by a Tobit procedure the log of exports which are censored at their minimum observed positive value to each destination country in the sample.46 45 Higher quality wines require higher quality inputs which can be imported. For instance, the oak barrels in which higher quality wines mature are typically imported from the US or Europe, the most prestigious barrels being made from French wood (Thornton, 2013). In addition to the effect that we identify in Table 10, we could therefore expect pass-through to be lower for the import-intensive exporters that rely heavily on imported intermediate inputs (Amiti, Itskhoki, and Konings, 2014), this effect being in turn weaker for the higher quality goods that involve higher quality imported inputs (Bernini and Tomasi, 2015). As firm-level imports are unavailable, we are unable to explore this channel. 46 Amiti et al. (2014) discuss the potential bias in pass-through regressions due to sample selection. Intuitively, we would expect the firms that drop out from the sample in response to an appreciation to suffer from an adverse productivity shock, and therefore to want to raise prices more than the firms staying in the sample. Censoring of these firms leads

18

ACCEPTED MANUSCRIPT

Local Distribution Costs

NU

4.2

SC

RI P

T

The results are reported in Table 11. The first two columns include positive trade flows only. Column (1) reproduces the first column of Table 10 for export volumes, while column (2) further includes the exchange rate interacted with quality. The exchange rate response of export volumes decreases with quality, and quality is negative. Columns (3) and (4) include the extensive margin, and report the corresponding Tobit estimates. Interestingly, the relationship between quality and exports is now positive, confirming that revenue increases with quality, and that a selection bias is present in the positive trade flows sample. Most importantly, our main variable of interest, i.e., the exchange rate interacted with quality, remains negative and significant in column (4), and its magnitude is similar to the one reported in column (2). As a result, in the remainder of the paper we focus on the intensive margin sample only as quality drops out once product-year fixed effects are controlled for.47

PT

ED

MA

Based on the assumption that higher quality goods have higher (per unit) distribution costs, our model predicts that the heterogeneous response of prices and quantities to changes in exchange rates due to differences in product quality should be more acute when distribution costs are higher. To explore the relevance of this proposition, we measure distribution costs using the cross-country data compiled by Anderson and Nelgen (2011) for three categories of wine products in 2009: super-premium (with a CIF import price above 7.5 US dollars per liter), commercial-premium (between 2.5 and 7.5 US dollars), and non-premium (below 2.5 US dollars). Following Anderson and Nelgen (2011), we assume that distribution costs for non-premium and premium wines are equal to 25 and 33 percent of their CIF import price augmented by import, excise, and value-added taxes.48 Across the countries included in our sample, we find that distribution costs are on average equal to 0.9, 1.4, and 2.7 US dollars per liter for non-premium, commercial-premium, and super-premium wines, respectively.

AC

CE

To merge the distribution costs with our exports data, one issue is that the three categories of wine products are defined by CIF import price while in the customs data set we only observe Argentinean FOB export prices. To match the two data sets we therefore calculate, using bilateral trade data for 2009 from Comtrade, the bilateral CIF over FOB ratio of still wine (HS codes 22.04.21 and 22.04.29) exported from Argentina to each destination country included in our sample. This allows us to convert the CIF import price thresholds of 2.5 and 7.5 US dollars into FOB export price thresholds that vary across destination countries, and to merge the distribution costs of the three wine categories with the wines included in our sample by FOB export unit value (in US dollars per liter).49 to an upward bias in the elasticity of export prices to the exchange rate (i.e., a downward bias in pass-through). This upward bias should however be less severe for the high performance firms that have a lower probability of exit, leading to underestimate the effect of firm-level performance on the exchange rate elasticity. In Panel A of Table 10, sample selection could therefore lead to an upward bias in the exchange rate elasticity, and to a downward bias in the exchange rate interacted with quality. 47 On the issue of selection, Campos (2010) argues that the intensive and extensive margins of adjustment have opposite effects on pass-through. In an online appendix available from our websites, we show that our results remain robust to restricting the sample to permanent exporters only. 48 The import prices, and import, excise, and value-added taxes for the three wine categories are reported for 44 countries and for seven regional groupings of other countries. When the data for a country are missing, if available we use instead the data for the regional grouping the country belongs to. Due to missing observations, our sample is slightly reduced. 49 Due to data constraints, this approach is subject to a number of limitations. First, the distribution costs and FOB price thresholds can only be measured for the year 2009. Second, the distribution costs for the three categories of wine products can only be calculated for total wine imports, and not for Argentinean imports only. Third, we cannot calculate the CIF over FOB ratio separately for each of the three wine categories.

19

ACCEPTED MANUSCRIPT

RI P

T

In column (1) of Panel A in Table 12 we regress equation (11) for unit values, but only interact the exchange rate with (log) distribution costs (product fixed effects interacted with dummies for non-premium, commercial-premium, and super-premium wines are also included). The positive and significant coefficient on the interaction term provides direct evidence that pricing-to-market increases with distribution costs. At the mean value of local costs, the exchange rate elasticity is equal to 0.115 (significant at the one percent level). This elasticity increases with distribution costs, and is equal to 0.088 at the 5th percentile of the distribution of local costs, and to 0.133 at the 95th percentile (both significant at the one percent level), which corresponds to a 51 percent increase in pricing-to-market.

ED

MA

NU

SC

Next, we divide our sample into two groups of high and low distribution costs according to whether these costs lie above or below their median in the sample. Quality varies between 55 and 97 when local costs are above the median, and between 55 and 96 below. Still, the Pearson correlation between distribution costs and quality is positive and equal to 15 percent.50 In columns (2) and (3) we then regress equation (11), separately for the high and low costs samples. The interaction between the exchange rate and quality is larger in magnitude when costs are high (at a value of 0.004) than when costs are low (at a value of 0.001 which is insignificant). In addition, at the mean value of quality, the exchange rate elasticity is equal to 0.151 when costs are high (significant at the one percent level), and to 0.093 when costs are low (significant at the five percent level). Therefore, firms price-to-market more, and by more for higher quality wines, when local costs are high.

AC

CE

PT

In column (4), we estimate equation (11) over the full sample, and include interactions between the exchange rate, quality, and local costs. At the mean value of local costs and of quality, the response of unit values to a change in the exchange rate is equal to 0.116 (significant at the one percent level). This elasticity increases with quality, and is equal to 0.110 at the 5th percentile of the quality distribution, and to 0.121 at the 95th percentile (both significant at the one percent level), which corresponds to a ten percent increase in pricing-to-market. Crucially, the triple interaction between the exchange rate, quality, and local costs is positive and significant. Therefore, the exchange rate response of unit values to quality increases with local costs. Overall, our findings are consistent with Berman et al. (2012), Chatterjee et al. (2013), and Li et al. (2015) who find that pass-through decreases with local costs, but contrary to Auer et al. (2014) who conclude that distribution costs do not explain the lower pass-through rate of higher quality cars.51 For export volumes, in column (1) of Panel B the interaction between the exchange rate and local costs is negative, as expected, but insignificant. Still, columns (2) and (3) show that in response to a real depreciation, exporters increase by less the exports of higher quality wines only when local costs are high. Most importantly, the triple interaction between the exchange rate, quality, and local costs in column (4) is negative and significant. In other words, the higher the costs, the smaller the exchange rate response of higher quality exports. 50

Also consistent with our model, local costs tend to be higher in higher income countries (the correlation with real GDP per capita is equal to 28 percent). 51 Also, Campa and Goldberg (2010) show that local distribution costs lower exchange rate pass-through into import prices. Hellerstein (2008) shows that partial pass-through can be explained by markup adjustments and by the presence of local costs in roughly similar proportions.

20

ACCEPTED MANUSCRIPT 4.3

Heterogeneity across Destination Countries

MA

NU

SC

RI P

T

We estimate equation (11) for unit values, and include interactions between the real exchange rate, quality, and real GDP per capita. The results are reported in column (1) of Panel A in Table 13. At the mean value of per capita income and of quality, the response of unit values to a change in the exchange rate is significant at the one percent level and is equal to 0.150. Consistent with Prediction 1, this elasticity increases with quality, and is equal to 0.134 and 0.168 (both significant at the one percent level) at the 5th and 95th percentiles of the quality distribution, respectively. Most importantly, the triple interaction between the exchange rate, quality, and real GDP per capita is positive and significant, which indicates that the exchange rate response of unit values to quality increases with per capita income, providing direct support to Prediction 3. This finding is consistent with the assumption of Alessandria and Kaboski (2011) and Simonovska (2015) that consumers in richer countries have lower demand elasticities than consumers in lower income countries, allowing firms to extract higher markups from richer destinations.52 It is also in line with Alessandria and Kaboski (2011) who show that US firms price-to-market more in richer destinations.

CE

PT

ED

According to our model of Section 2, the exchange rate elasticity of unit values decreases with the trade costs of exporting to each destination. We therefore check that our findings are not driven by differences in trade costs between high and low income countries. Column (2) further interacts the exchange rate with the bilateral distance between Argentina and each destination country. Consistent with the model, the interaction is negative and significant. Most importantly, controlling for bilateral distance does not alter the evidence in favor of Prediction 3 as the triple interaction remains positive and significant. At the mean value of quality, per capita income, and of bilateral distance, the exchange rate elasticity is equal to 0.127 (significant at the one percent level). It is equal to 0.110 and 0.144 (significant at the ten and at the one percent levels) at the 5th and 95th percentiles of the quality distribution, respectively.

AC

Also predicted by our model is that pricing-to-market increases with local distribution costs in the importing country. As the latter are correlated with per capita income, we verify that our results are not driven by cross-country differences in distribution costs, and include in column (3) an interaction between the exchange rate and distribution costs. Again, this exercise does not qualitatively affect the evidence in favor of Prediction 3 as the triple interaction remains positive and significant. The exchange rate elasticity evaluated at the mean value of quality, per capita income, and of distribution costs is equal to 0.106, and rises from 0.096 to 0.115 from the 5th to the 95th percentile of the quality distribution (all elasticities being significant at the one percent level). As the Wine Spectator is a US-based rating, a final concern is that it may not be representative of taste preferences for quality in other countries.53 In column (4), we therefore drop the US from the sample. Finally, in column (5), we simultaneously drop the US and control for the effects of distance 52 It however contrasts with Crozet et al. (2012) who find that French exporters of Champagne charge higher prices to lower income countries. Their interpretation is that lower income countries do not produce much sparkling wine, therefore French exporters charge higher prices when they enter the markets of these countries. We checked, and confirm, that our results remain robust to controlling for the competition that may arise from wine production in destination countries. We estimated equations (11) and (12) and included total wine output (in tons) of each destination country in each year (from the Food and Agriculture Organization of the United Nations), on its own and interacted with the exchange rate. 53 For instance, US consumers have a preference for fruiter wines with less alcoholic content, while Europeans prefer less fruity wines with higher alcohol content (Artopoulos, Friel, and Hallak, 2011).

21

ACCEPTED MANUSCRIPT

T

and of distribution costs on the response of unit values to exchange rate changes. In both cases, the evidence in favor of Prediction 3 continues to hold.54 The results for export volumes are reported in Panel B. Consistent with Prediction 4, the triple interactions between the exchange rate, quality, and per capita income are negative, but insignificant.

Endogeneity

CE

5

PT

ED

MA

NU

SC

RI P

Another way to investigate our predictions is to split the countries included in our sample between high and low income, using the World Bank’s classification based on GNI per capita in 2011.55 We estimate equations (11) and (12), and interact the exchange rate, and its interaction with quality, with a dummy for high and a dummy for low income countries. The results for unit values are reported in column (1) of Panel A in Table 14. Interestingly, the coefficient on the exchange rate interacted with quality is positive and significant for higher income destinations only. This finding is consistent with Prediction 3, and with Auer et al. (2014) who predict that pass-through decreases with quality in higher income countries only (we do not find evidence, as predicted by Auer et al., 2014, that passthrough increases with quality in lower income countries). Qualitatively, the results remain robust to controlling for bilateral distance (column 2), distribution costs (column 3), excluding the US (column 4), and to all of them simultaneously (column 5). As in Alessandria and Kaboski (2011), pricing-tomarket is stronger for exports to richer destinations. For instance, in column (5), at the mean value of quality, distance, and of distribution costs, the exchange rate elasticity is insignificant for lower income countries. It is instead significant and equal to 0.106 for higher income countries, and increases from 0.099 to 0.114 from the 5th to the 95th percentile of the quality distribution (all significant at the one percent level). In Panel B, the results for export volumes provide direct support for Prediction 4 as in all cases, the interactions between the exchange rate and quality are significantly negative for higher income countries only.

AC

One concern with our estimations is the potential endogeneity of quality in explaining unit values and export volumes. First, although the Wine Spectator ratings are produced from blind tastings where the “price is not taken into account in scoring,” the “tasters are told [...] the general type of wine (varietal and/or region) and the vintage” year.56 Similarly for Parker, “neither the price nor the reputation of the producer/grower affect the rating in any manner,” but the “tastings are done in peer-group, single-blind conditions (meaning that the same types of wines are tasted against each other, and the producers names are not known).”57 As a result, we cannot rule out that the basic information that the experts hold about each wine, in combination with their own subjective preferences, affects in a way or the other their scores, leading to an endogeneity bias which direction is, however, unclear. Second, 54 Some papers argue that the demand for quality is not only determined by income, but also by the distribution of income within a country, and that richer households typically spend a larger share of their income on higher quality goods. Fajgelbaum et al. (2011) show that richer, as well as more unequal countries, have a larger demand for higher quality goods. Mitra and Trindade (2005) show that more unequal countries import more of higher quality goods. If the demand for quality is stronger for more unequal countries, the response of unit values to exchange rate changes should increase with quality by more for more unequal countries. We estimated equation (11) and included interactions between the exchange rate, quality, and the Gini coefficient of each destination (from the World Bank World Development Indicators), while controlling for per capita income. The exchange rate response of unit values to quality increases by more for more equal than for more unequal countries, where the distribution of income tends to be more equal in higher income countries. 55 Low income countries have a GNI per capita of less than 4,035 US dollars. 56 See http://www.winespectator.com/display/show?id=about-our-tastings. 57 See https://www.erobertparker.com/info/legend.asp.

22

ACCEPTED MANUSCRIPT

RI P

T

some papers question the reliability and usefulness of wine ratings as a measure of quality. Ashenfelter and Quandt (1999) and Quandt (2007) argue that wine ratings fail to convey any useful information about quality, and observe that the disagreements between experts are substantial.58 Hodgson (2008) finds that experts are also inconsistent as they do not assign the same scores to the same wines at different tastings. Measurement error in the quality scores can thus generate a positive endogeneity bias as wine producers are likely to charge a higher price for a wine that is awarded a higher score because they believe consumers are willing to pay more for higher ranked wines. To tackle the endogeneity of quality, we use appropriate instruments.59

PT

ED

MA

NU

SC

The instruments we rely on to explain quality include geography and weather-related factors. Indeed, the literature argues that the amount of rainfall, and the average temperature during the growing season, are strong determinants of wine quality (Ashenfelter, 2008; Thornton, 2013). In the Southern hemisphere, the growing season spans the period from September (in the year before the vintage year) to March. In order to allow the effects of temperature and of rainfall to be nonlinear throughout the growing season, we consider as instruments the average temperature, and the total amount of rainfall, in each growing province in each month between September and March.60 Besides, one particularity of Argentina’s wine industry is the high altitude at which some of the growing regions are located, which reduces the problems related to insects or grape diseases that affect quality at a low altitude. We therefore use the altitude of each province as an additional instrument for quality. The data on monthly average temperature (in degrees Celsius), total rainfall (in millimeters), and altitude (in meters) are from the National Climatic Data Center of the US Department of Commerce.61 Gaps in the data are filled using online information, although missing information for some provinces and vintage years results in a slightly reduced sample.62

AC

CE

As the instruments are only available over a reduced sample, we first replicate our benchmark OLS estimations reported in column (3) of Table 10. Next, we regress by Instrumental Variables (IV) unit values and export volumes on the exchange rate and its interaction with quality. The instruments for the interaction term include the monthly temperature, rainfall, and altitude variables, each interacted with the exchange rate. As our instruments could be correlated with unobserved productivity shocks that may vary across export destinations (e.g., Chaney, 2008), we also report specifications with firmdestination-year fixed effects to ensure that the exclusion restriction of our instruments is satisfied. Panel A in Table 15 reports the OLS and IV specifications for unit values. The OLS results in column (1) show that qualitatively, our main findings go through over the smaller sample. As can be seen in column (2), the instrumentation inflates the coefficient on the interaction term at a value of 0.015 compared to an OLS estimate of 0.001 in column (1). The OLS and IV exchange rate elasticities 58 Psychologists argue that the ratings are subjective and depend on the tasters’ personal preferences. For instance, Parker is known to have a preference for fruit-flavored wines, and rewards them with a higher score (Thornton, 2013). 59 Endogeneity could also arise due to measurement error in the real exchange rate, and in particular in the Argentinean CPI. Since 2007, the credibility of the official CPI data has indeed been widely questioned by the public. The IMF has issued a declaration of censure, and called on Argentina to adopt remedial measures to address the quality of the official CPI. Further information can be found at http://www.imf.org/external/np/sec/pr/2013/pr1333.htm. In our regressions, the product-year fixed effects control for measurement error in the Argentinean CPI. 60 The temperatures, rainfall, and altitudes are measured at the following weather stations (provinces): Catamarca Aero (Catamarca), Paran´ a Aero (Entre R´ıos), Santa Rosa Aero (La Pampa), La Rioja Aero (La Rioja), Mendoza Aero (Mendoza), Neuqu´en Aero (Neuquen), Bariloche Aero (R´ıo Negro), Salta Aero (Salta), and San Juan Aero (San Juan). 61 This is available at http://www.ncdc.noaa.gov/IPS/mcdw/mcdw.html. 62 Online information is taken from www.tutiempo.net.

23

ACCEPTED MANUSCRIPT

RI P

T

are respectively equal to 0.226 and 0.174 at the 5th percentile of the quality distribution, and increase to 0.244 and 0.360 at the 95th percentile of the quality distribution (all significant at the one percent level). In other words, the instrumentation decreases exchange rate pass-through for higher quality wines. For export volumes in Panel B, the IV coefficient on the interaction term is insignificant. For the IV regressions, the Kleibergen-Paap F statistic (equal to 93, with a critical value equal to 21, Stock and Yogo, 2005) rejects the null of weak correlation between the excluded instruments and the endogenous regressors.

Extensions

MA

6

NU

SC

Columns (3) and (4) report the OLS and IV estimates obtained with firm-destination-year fixed effects. Again, the IV estimates are larger in magnitude compared to the OLS coefficients, and they are significant for both unit values and export volumes in Panels A and B, respectively. The firststage estimates are reported in an online appendix available from our websites, and show that climate variation and geography affect the interaction between the exchange rate and quality.

6.1

ED

This section discusses extensions to our benchmark results. It shows that the effect of quality on pass-through is nonlinear, and that pricing-to-market increases with firm-level productivity, exporter quality, and export market shares.

Nonlinearities

CE

PT

Under the assumption that higher quality wines have higher (per unit) distribution costs, the elasticity of prices to the exchange rate should be larger for higher quality wines. We therefore estimate equation (11), allowing for nonlinearities in the effect of quality by letting the interaction between the exchange rate and quality vary between the six quality bins defined in Table 3.

AC

The results are reported in Table 16. Column (1) shows that the exchange rate interacted with quality is about three times larger for “Great” (with a score above 94) than for lower ranked wines. As we cannot reject that the coefficients on the interactions for the lower quality wines are equal, in column (2) we classify wines into two different bins only, i.e., with a score above or below 94. The coefficient on the interaction term is still much larger for “Great” wines, consistent with higher quality goods having higher (per unit) distribution costs. In column (2), the exchange rate elasticities evaluated at the mean value of quality within each of the two quality bins show that in response to a ten percent real depreciation, exporters raise their peso prices by 1.9 percent for the lowest quality wines (at a mean quality of 85, significant at the one percent level), and by 6.4 percent for the top quality wines (at a mean quality of 96, significant at the five percent level). The results for volumes are reported in columns (3) and (4). In column (4), the exchange rate interacted with quality is significantly negative for the lower ranked wines only.

6.2

Productivity

As the literature has focused on firm-level productivity in driving heterogeneous pass-through (e.g., Berman et al., 2012; Chatterjee et al., 2013), we check whether more productive firms price-to-market more than others. 24

ACCEPTED MANUSCRIPT

The Quality of Exporters

MA

6.3

NU

SC

RI P

T

Without any information on the firm-level characteristics that are required to measure productivity (such as value-added or employment), we construct a variable that proxies for firm size as the latter has been shown to correlate strongly with productivity (Bernard, Eaton, Jensen, and Kortum, 2003). We use the total volume (in liters) of FOB exports of each firm in each year, and divide our sample into two groups of large and small firms, according to whether their total yearly exports are above or below median total exports in the sample. Columns (1) and (2) of Table 17 report the results for unit values and export volumes, separately for large and small firms. Panel A shows that only the larger, and therefore the more productive firms, price-to-market (at the mean value of quality, the elasticity is equal to 0.164 and is significant at the one percent level in column 1, but is insignificant in column 2), and by more for higher quality wines. Symmetrically, Panel B reveals that, in response to a real depreciation, smaller firms increase volumes by more than others (at the mean value of quality, the exchange rate elasticity is equal to 2.762 and is significant at the five percent level, but is insignificant for the larger firms).

PT

ED

As illustrated by Table 7, firms export wines with heterogeneous levels of quality. Table 9 shows there is also some heterogeneity in the average quality exported by firms. Firms can therefore be divided into two groups of higher and lower quality exporters by splitting the sample at the median of average firm-level quality, which varies between 72 and 92. This sample split is not very precise as average firmlevel quality excludes unrated wines, but it still captures that firms export wines with heterogeneous levels of quality: quality varies between 55 and 97 for higher quality exporters (average quality is equal to 87), and between 55 and 94 for lower quality exporters (average quality is equal to 83).

AC

CE

Columns (3) and (4) of Table 17 report the results for unit values and export volumes, separately for the two groups of exporters. At the mean value of quality, pricing-to-market is stronger for higher than for lower quality exporters (the elasticities are equal to 0.258 and 0.127 and are significant at the one and five percent levels, respectively). Pricing-to-market increases with quality for higher quality exporters only, while the response of export volumes to a change in the exchange rate decreases with quality for both higher and lower quality exporters.

6.4

Export Market Shares

According to Amiti et al. (2014) and Atkeson and Burstein (2008), exporters have lower markups in the markets where they have a smaller market share, making it difficult to adjust their markups in response to exchange rate changes. As a result, firms price-to-market less, and have a higher passthrough, in the countries where they own a small share of the market. To check if this holds in our data, we compute the market share of each firm by export destination and by year as the share of its exports in the total value exported by all firms by grape, type, province of origin, and vintage year. Based on the median market share, we split the sample between high and low market shares. Quality varies between 55 and 97 when market shares are high, and between 55 and 96 when they are low. In columns (1) and (2) of Panel A in Table 18, we regress equation (11) for unit values, separately for the two samples. Consistent with expectations, pass-through is higher when market shares are low (at the mean value of quality, the exchange rate elasticity is equal to 0.190 when market shares are high, and to 0.143 when they are low, significant at the one and ten percent levels, respectively). Also, 25

ACCEPTED MANUSCRIPT pass-through falls with quality only when market shares are high. Column (3) only controls for the effect of market shares in the full sample, and confirms that pass-through falls with market shares.

NU

SC

RI P

T

In the models of Amiti et al. (2014) and Atkeson and Burstein (2008), quality affects pass-through only through market shares. Therefore, quality should stop having predictive power once market shares are controlled for. In column (4), we estimate equation (11) over the full sample, and separately interact the exchange rate with quality and market shares. Interestingly, the two interactions are positive and significant, implying that quality matters beyond the effect of market shares. At the mean value of market shares, the exchange rate elasticities at the 5th and 95th percentiles of the quality distribution are equal to 0.182 and 0.201 (both significant at the one percent level), or an 11 percent difference in pass-through. At the mean value of quality, the elasticities at the 5th and 95th percentiles of the market shares distribution are equal to 0.186 and 0.195 (both significant at the one percent level), or a five percent difference in pass-through. The contribution of market shares in explaining pass-through is thus smaller than that of quality, and is smaller than in Amiti et al. (2014).

Robustness

AC

7

CE

PT

ED

MA

Panel B reports the results for volumes. Columns (1) and (2) show that a real depreciation increases by less the exports of higher quality wines only when market shares are high, and has no effect when market shares are low. Therefore, and as confirmed by columns (3) and (4), a real depreciation benefits more the high market share firms. One possible interpretation is that market shares signal quality (Caminal and Vives, 1996). Following a real depreciation, the high market share firms expand at the expense of the others because consumers in the importing country believe that these firms are of higher quality. Another interpretation is that the high market share firms have acquired a good reputation in the destination market, which is an important determinant of success in export markets (Macchiavello, 2010). Finally, column (4) provides additional evidence that the effect of quality is not strictly captured by market shares as the exchange rate elasticity rises with market shares, but decreases with quality.63

This section discusses alternative specifications we implement to ensure the robustness of our findings. Despite some variation across specifications in the magnitude of the effect of quality on pass-through, the broad similarity of the results supports the paper’s main conclusions. To conserve space, the results of the exercises described in this section are reported in an online appendix available from our websites. On the measurement of quality, we first use a variable which takes on values between one and six, where each value corresponds to one of the different bins of the Wine Spectator. Second, to control for the fact that different tasters disagree on wine quality, we restrict the sample to the wines which Wine Spectator scores belong to the same quality bin as the Parker ratings. Third, we exclude the US from the sample as both the Wine Spectator and Parker ratings are US-based and therefore may not be representative of perceived wine quality in other countries. Fourth, we measure quality using the information provided by unit values. For each firm in each year, we rank wines by mean unit value in decreasing order, and use this rank as an inverted measure of quality (Berman et al., 2012; Chatterjee 63 To summarize, larger firms, with higher market shares and exporting a higher quality, price-to-market more than others, and by more for higher quality goods. Larger firms tend to have higher market shares (the correlation is 0.19). The correlations between firm-level quality, on the one hand, and market shares and firm size, on the other hand, are only equal to -0.12 and -0.04, respectively.

26

ACCEPTED MANUSCRIPT

T

et al., 2013; Mayer et al., 2014). Finally, to include unrated wines in the sample, we calculate an average Wine Spectator rating by wine name and type. We also identify the wines which vintage year is missing, assign to each of them the scores of the wines with the same name, grape, and type on a (1,6) scale, and assign a value of one to the wines produced by firms that only export unrated wines under the assumption that these firms only produce low quality wines (Crozet et al., 2012).

NU

SC

RI P

We provide robustness checks on the real exchange rate. First, as the peso was allowed to fluctuate with respect to the US dollar since 2002 within a “crawling band that is narrower than or equal to +/-2 percent” (Reinhart and Rogoff, 2004), variations in the US dollar to peso real exchange rate may have essentially come from movements in prices (Li et al., 2015). We therefore exclude from the sample the US and all the countries which currencies are pegged to the US dollar. Second, we include the nominal exchange rate, its interaction with quality, and separately control for the destination country’s CPI.

AC

CE

PT

ED

MA

To control for the exhaustible nature of wine, and the fact that once a wine with a specific vintage year runs out the producer can no longer produce, and therefore export that variety, we construct a new sample and define a product according to the name of the wine, its grape, and type, but ignore the vintage year.64 Also, as higher quality wines are subject to capacity constraints in production (Thornton, 2013), we control for total export volume by wine as a proxy for total output. To introduce dynamics, we include one lag, and then two lags, on the exchange rate and its interaction with quality, and to address non-stationarity we estimate equations in first differences. Our results hold when controlling for GDP or GDP per capita in the pass-through regressions, restricting the sample to the post–2002 and pre–2008 periods, interacting the real exchange rate with year dummies, estimating our equations at a quarterly frequency, and including in the sample wholesalers and retailers or the permanent exporters only. They also remain robust to including firm-destination-year fixed effects which control for excluded characteristics that vary by firm-destination-year (such as the time-varying demand of a country for a firm’s exports, or the presence of long term contracts between exporters and importers in each destination). Lastly, although we do not observe the currency in which Argentinean firms price their exports, we report estimates using unit values denominated in US dollars per liter.

8

Concluding Remarks

This paper analyzes the heterogeneous response of exporters to changes in real exchange rates due to differences in product quality. To understand the mechanism through which quality affects the pricing and quantity decisions of firms, the first contribution of the paper is to present a model where firms export multiple products with heterogeneous levels of quality, and where the demand elasticity falls with quality. The second contribution is to bring the testable predictions of the model to the data. We combine a unique data set of Argentinean firm-level destination-specific export values and volumes of highly disaggregated wine products between 2002 and 2009 with a data set on experts wine ratings to measure quality. Given the very rich nature of our data set, our main sample includes 6,720 different wines exported by 209 wine producers over the period. Overall, our results find strong support for the predictions of the model, and quantitatively, our benchmark regression shows that pricing-to-market is 10.3 percent higher for a wine at the 95th than at the 5th percentile of the quality distribution. 64

This exercise allows us to get closer to the CES utility function assumption of a fixed set of varieties as there is no reason for wines to run out once we exclude the vintage year.

27

ACCEPTED MANUSCRIPT

NU

SC

RI P

T

By showing that pass-through decreases with quality, our empirical analysis validates different classes of models based on endogenous and variable markups, including Atkeson and Burstein (2008), Auer et al. (2014), Corsetti and Dedola (2005), and Melitz and Ottaviano (2008). As in Berman et al. (2012), our results are therefore supportive of models with variable demand elasticities which are able to generate heterogeneous pricing-to-market. We do not find evidence, however, that quality affects passthrough only through market shares, as implied by Atkeson and Burstein (2008), or that pass-through rises with quality in lower income countries, as predicted by Auer et al. (2014). Missing any data on consumer prices, we are unable to check whether the demand elasticity to the consumer price decreases or not with quality, which would allow us to distinguish between the Corsetti and Dedola (2005) and Melitz and Ottaviano (2008) models. However, as we find evidence that local distribution costs reduce pass-through, especially for higher quality wines, we conclude that our extension to Corsetti and Dedola (2005) does a good job at explaining why pass-through decreases with quality.

ED

MA

One promising avenue for future research would be to use our data to identify the pricing-tomarket behavior of exporters by assuming that differences in the prices of the same wines sold to multiple destinations reflect differences in markups as these wines are subject to the same marginal cost. This approach has been adopted by Auer et al. (2014), Burstein and Jaimovich (2012), Cavallo et al. (2014), Fitzgerald and Haller (2014), and Simonovska (2015).

References

PT

Alessandria, G., Kaboski, J.P., 2011. Pricing-to-market and the failure of absolute PPP. American Economic Journal: Macroeconomics 3 (1), 91–127.

CE

Amiti, M., Itskhoki, O., Konings, J., 2014. Importers, exporters, and exchange rate disconnect. American Economic Review 104 (7), 1942–1978.

AC

Amiti, M., Khandelwal, A.K., 2013. Import competition and quality upgrading. Review of Economics and Statistics 95 (2), 476–490. Anderson, K., Nelgen, S., 2011. Global Wine Markets, 1961–2009: A Statistical Compendium. University of Adelaide Press, Adelaide. Antoniades, A., Zaniboni, N., 2015. Exchange rate pass-through into retail prices. International Economic Review, forthcoming. Artopoulos, A., Friel, D., Hallak, J.C., 2011. Lifting the domestic veil: The challenges of exporting differentiated goods across the development divide. NBER Working Paper 16947. Ashenfelter, O., Quandt, R., 1999. Analyzing a wine tasting statistically. Chance 12 (3), 16–20. Ashenfelter, O., 2008. Predicting the quality and prices of Bordeaux wine. Economic Journal 118 (529), F174–F184. Atkeson, A., Burstein, A., 2008. Trade costs, pricing-to-market, and international relative prices. American Economic Review 98 (5), 1998–2031. Auer, R.A., Chaney, T., 2009. Exchange rate pass-through in a competitive model of pricing-to-market. Journal of Money, Credit and Banking 41 (1), 151–175. Auer, R.A., Chaney, T., Saur´e, P., 2014. Quality pricing-to-market. CEPR Discussion Paper 10053. Baldwin, R.E., Harrigan, J., 2011. Zeros, quality, and space: Trade theory and trade evidence. American Economic Journal: Microeconomics 3 (2), 60–88. 28

ACCEPTED MANUSCRIPT Basile, R., de Nardis, S., Girardi, A., 2012. Pricing to market, firm heterogeneity and the role of quality. Review of World Economics 148 (4), 595–615. Bastos, P., Silva, J., 2010. The quality of a firm’s exports: Where you export to matters. Journal of International Economics 82 (2), 99–111.

RI P

T

Bastos, P., Silva, J., Verhoogen, E., 2014. Export destinations and input prices. NBER Working Paper 20143. Bellone, F., Musso, P., Nesta, L., Warzynski, F., 2016. International trade and firm-level markups when location and quality matter. Journal of Economic Geography 16 (1), 67–91.

SC

Berman, N., Martin, P., Mayer, T., 2012. How do different exporters react to exchange rate changes? Quarterly Journal of Economics 127 (1), 437–492.

NU

Bernard, A.B., Eaton, J., Jensen, J.B., Kortum, S., 2003. Plants and productivity in international trade. American Economic Review 93 (4), 1268–1290. Bernini, M., Tomasi, C., 2015. Exchange rate pass-through and product heterogeneity: Does quality matter on the import side? European Economic Review 77, 117–138.

MA

Berthou, A., Emlinger, C., 2011. The trade unit values database. CEPII Working Paper 2011–10. Burstein, A., Gopinath, G., 2014. International prices and exchange rates, in: Gopinath, G., Helpman, E., Rogoff, K. (Eds), Handbook of International Economics, Vol. 4., Elsevier, Amsterdam, pp. 391–451.

ED

Burstein, A., Jaimovich, N., 2012. Understanding movements in aggregate and product-level real exchange rates. University of California, Los Angeles, mimeo.

PT

Burstein, A., Neves, J.C., Rebelo, S., 2003. Distribution costs and real exchange rate dynamics during exchange-rate-based stabilizations. Journal of Monetary Economics 50 (6), 1189–1214. Caminal, R., Vives, X., 1996. Why market shares matter: An information-based theory. RAND Journal of Economics 27 (2), 221–239.

CE

Campa, J.M., Goldberg, L.S., 2005. Exchange-rate pass-through into import prices. Review of Economics and Statistics 87 (4), 679–690.

AC

Campa, J.M., Goldberg, L.S., 2010. The sensitivity of the CPI to exchange rates: Distribution margins, imported inputs, and trade exposure. Review of Economics and Statistics 92 (2), 392–407. Campos, C.F.S., 2010. Incomplete exchange rate pass-through and extensive margin of adjustment. Yale University, mimeo. Cavallo, A., Neiman, B., Rigobon, R., 2014. Currency unions, product introductions, and the real exchange rate. Quarterly Journal of Economics 129 (2), 529–595. Chaney, T., 2008. Distorted gravity: The intensive and extensive margins of international trade. American Economic Review 98 (4), 1707–1721. Chatterjee, A., Dix-Carneiro, R., Vichyanond, J., 2013. Multi-product firms and exchange rate fluctuations. American Economic Journal: Economic Policy 5 (2), 77–110. Chen, N., Juvenal, L., 2013. Quality, trade, and exchange rate pass-through. CEPR Discussion Paper 9744. Corsetti, G., Dedola, L., 2005. A macroeconomic model of international price discrimination. Journal of International Economics 67 (1), 129–155. Crin`o, R., Epifani, P., 2012. Productivity, quality, and export behaviour. Economic Journal 122 (565), 1206–1243. Crozet, M., Head, K., Mayer, T., 2012. Quality sorting and trade: Firm-level evidence for French wine. Review of Economic Studies 79 (2), 609–644. 29

ACCEPTED MANUSCRIPT De Loecker, J., Warzynski, F., 2012. Markups and firm-level export status. American Economic Review 102 (6), 2437–2471. Dixit, A.K., Stiglitz, J.E., 1977. Monopolistic competition and optimum product diversity. American Economic Review 67 (3), 297–308.

RI P

T

Dornbusch, R., 1989. Real exchange rates and macroeconomics: A selective survey. Scandinavian Journal of Economics 91 (2), 401–432. Engel, C., 2002. Expenditure switching and exchange rate policy. NBER Macroeconomics Annual 17, 231–272.

SC

Fajgelbaum, P., Grossman, G.M., Helpman, E., 2011. Income distribution, product quality, and international trade. Journal of Political Economy 119 (4), 721–765.

NU

Fan, H., Li, Y.A., Yeaple, S.R., 2014. Trade liberalization, quality, and export prices. NBER Working Paper 20323. Feenstra, R.C., Levinsohn, J.A., 1995. Estimating markups and market conduct with multidimensional product attributes. Review of Economic Studies 62 (1), 19–52.

MA

Feenstra, R.C., Romalis, J., 2014. International prices and endogenous quality. Quarterly Journal of Economics 129 (2), 477–527.

ED

Fitzgerald, D., Haller, S., 2014. Pricing-to-market: Evidence from plant-level prices. Review of Economic Studies 81 (2), 761–786. Gagnon, E., Mandel, B.R., Vigfusson, R.J., 2014. Missing import price changes and low exchange rate pass-through. American Economic Journal: Macroeconomics 6 (2), 156–206.

PT

Goldberg, P.K., Knetter, M.M., 1997. Goods prices and exchange rates: What have we learned? Journal of Economic Literature 35 (3), 1243–1272.

CE

Goldberg, P.K., Verboven, F., 2001. The evolution of price dispersion in the European car market. Review of Economic Studies 68 (4), 811–848.

AC

Gopinath, G., Itskhoki, O., 2010. Frequency of price adjustment and pass-through. Quarterly Journal of Economics 125 (2), 675–727. Gopinath, G., Itskhoki, O., Rigobon, R., 2010. Currency choice and exchange rate pass-through. American Economic Review 100 (1), 304–336. Gopinath, G., Rigobon, R., 2008. Sticky borders. Quarterly Journal of Economics 123 (2), 531–575. Hallak, J.C., 2006. Product quality and the direction of trade. Journal of International Economics 68 (1), 238–265. Hallak, J.C., Schott, P.K., 2011. Estimating cross-country differences in product quality. Quarterly Journal of Economics 126 (1), 417–474. Hallak, J.C., Sivadasan, J., 2013. Product and process productivity: Implications for quality choice and conditional exporter premia. Journal of International Economics 91 (1), 53–67. Haltmaier, J., 2011. Empirical estimation of trend and cyclical export elasticities. Board of Governors of the Federal Reserve System, International Finance Discussion Paper 1030. Hellerstein, R., 2008. Who bears the cost of a change in the exchange rate? Pass-through accounting in the case of beer. Journal of International Economics 76 (1), 14–32. Helpman, E., Melitz, M., Rubinstein, Y., 2008. Estimating trade flows: Trading partners and trading volumes. Quarterly Journal of Economics 123 (2), 441–487. Hodgson, R.T., 2008. An examination of judge reliability at a major US wine competition. Journal of Wine Economics 3 (2), 105–113. 30

ACCEPTED MANUSCRIPT Hummels, D., 2007. Transportation costs and international trade in the second era of globalization. Journal of Economic Perspectives 21 (3), 131–154. Hummels, D., Klenow, P.J., 2005. The variety and quality of a nation’s exports. American Economic Review 95 (3), 704–723.

RI P

T

Hummels, D., Skiba, A., 2004. Shipping the good apples out? An empirical confirmation of the AlchianAllen conjecture. Journal of Political Economy 112 (6), 1384–1402. Johnson, R.C., 2012. Trade and prices with heterogeneous firms. Journal of International Economics 86 (1), 43–56.

SC

Joseph, R., 2012. Lifting the stone on the UK wine trade. http://thejosephreport.blogspot.fr/2012/08/lifting-stone-on-uk-wine-trade.html.

NU

Khandelwal, A., 2010. The long and short (of) quality ladders. Review of Economic Studies 77 (4), 1450–1476. Khandelwal, A., Schott, P.K., Wei, S.-J., 2013. Trade liberalization and embedded institutional reform: Evidence from Chinese exporters. American Economic Review 103 (6), 2169–2195.

MA

Knetter, M.M., 1989. Price discrimination by US and German exporters. American Economic Review 79 (1), 198–210. Knetter, M.M., 1992. Exchange rates and corporate pricing strategies. NBER Working Paper 4151.

ED

Knetter, M.M., 1993. International comparisons of pricing-to-market behavior. American Economic Review 83 (3), 473–486.

PT

Kugler, M., Verhoogen, E., 2012. Prices, plant size, and product quality. Review of Economic Studies 79 (1), 307–339. Li, H., Ma, H., Xu, Y., 2015. How do exchange rate movements affect Chinese exports? A firm-level investigation. Journal of International Economics 97 (1), 148–161.

CE

Macchiavello, R., 2010. Development uncorked: Reputation acquisition in the new market for Chilean wines in the UK. University of Warwick, mimeo.

AC

Manova, K., Zhang, Z., 2012a. Export prices across firms and destinations. Quarterly Journal of Economics 127 (1), 379–436. Manova, K., Zhang, Z., 2012b. Multi-product firms and product quality. NBER Working Paper 18637. Marston, R.C., 1990. Pricing to market in Japanese manufacturing. Journal of International Economics 29 (3/4), 217–236. Mayer, T., Melitz, M.J., Ottaviano, G.I.P., 2014. Market size, competition, and the product mix of exporters. American Economic Review 104 (2), 495–536. Melitz, M.J., 2003. The impact of trade on intra-industry reallocations and aggregate industry productivity. Econometrica 71 (6), 1695–1725. Melitz, M.J., Ottaviano, G.I.P., 2008. Market size, trade, and productivity. Review of Economic Studies 75 (1), 295–316. Mitra, D., Trindade, V., 2005. Inequality and trade. Canadian Journal of Economics 38 (4), 1253–1271. Nakamura, E., Steinsson, J., 2012. Lost in transit: Product replacement bias and pricing to market. American Economic Review 102 (7), 3277–3316. Nakamura, E., Zerom, D., 2010. Accounting for incomplete pass-through. Review of Economic Studies 77 (3), 1192–1230. Phillips, L.W., Chang, D.R., Buzzell, R.D., 1983. Product quality, cost position and business performance: A test of some key hypotheses. Journal of Marketing 47 (2), 26–43. 31

ACCEPTED MANUSCRIPT Quandt, R.E., 2007. On wine bullshit: Some new software? Journal of Wine Economics 2 (2), 129–135. Reinhart, C.M., Rogoff, K.S., 2004. The modern history of exchange rate arrangements: A reinterpretation. Quarterly Journal of Economics 119 (1), 1–48.

T

Rodr´ıguez-L´opez, J.A., 2011. Prices and exchange rates: A theory of disconnect. Review of Economic Studies 78 (3), 1135–1177.

RI P

Schott, P.K., 2004. Across-product versus within-product specialization in international trade. Quarterly Journal of Economics 119 (2), 647–678.

SC

Simonovska, I., 2015. Income differences and prices of tradables: Insights from an online retailer. Review of Economic Studies 82 (4), 1612–1656.

NU

Stock, J.H., Yogo, M., 2005. Testing for weak instruments in linear IV regression, in: Andrews, D.W.K., Stock, J.H. (Eds), Identification and Inference for Econometric Models: Essays in Honor of Thomas Rothenberg. Cambridge University Press, Cambridge, pp. 80–108. Sutton, J., 1998. Technology and Market Structure: Theory and History. MIT Press, Cambridge.

MA

Thornton, J., 2013. American Wine Economics. An Exploration of the US Wine Industry. University of California Press, Berkeley and Los Angeles, California. Verhoogen, E., 2008. Trade, quality upgrading and wage inequality in the Mexican manufacturing sector. Quarterly Journal of Economics 123 (2), 489–530.

AC

CE

PT

ED

Yu, Z., 2013. Exchange rate pass-through, firm heterogeneity and product quality: A theoretical analysis. Federal Reserve of Dallas Globalization and Monetary Policy Institute Working Paper 141.

32

ACCEPTED MANUSCRIPT

Table 1: Price Breakdown of Non-EU Wine Sold in Retail Outlets in the UK 5.76

7.19

8.83

10.09

VAT Retail margin Duty Distributor margin Common Customs Tariff Transport Winemaker

0.96 1.92 1.90 0.11 0.11 0.13 0.63

1.20 2.40 1.90 0.21 0.11 0.13 1.25

1.47 2.94 1.90 0.40 0.11 0.13 1.88

1.68 3.36 1.90 0.51 0.11 0.13 2.40

SC

RI P

T

Retail price (pound sterling per bottle)

Notes: All figures are in pound sterling. UK taxes include a duty of 1.90 pound sterling per bottle and a VAT sales tax of

NU

20 percent which may vary depending on alcohol content. Non-EU wines are subject to a Common Customs Tariff which

MA

does not apply to wines from the EU. Source: Joseph (2012).

Table 2: Harmonized System Classification 12-digit

Description

22.04.10

22.04.10.10.000.D 22.04.10.90.000.G 22.04.10.90.100.M 22.04.10.90.900.F 22.04.21.00.100.A 22.04.21.00.200.F 22.04.21.00.900.U 22.04.29.00.100.W 22.04.29.00.200.B 22.04.29.00.900.P 22.04.30.00.000.X

Sparkling wine – Champagne variety Sparkling wine – Not Champagne variety Sparkling wine – Gassified wine (i.e., aerated using CO2) Sparkling wine – Other Sweet wine; ¡ 2 liters Fine wine; ¡ 2 liters Other wine; ¡ 2 liters Sweet wine; ¿ 2 liters Fine wine; ¿ 2 liters Other wine; ¿ 2 liters Other grape must

22.04.30

PT

CE

22.04.29

AC

22.04.21

ED

6-digit

Notes: The table reports the 6-digit HS codes for wine products, the corresponding 12-digit codes as used by the Argentinean customs, and a description of the wines included under each code.

Table 3: Experts Ratings Wine Spectator (50,100) 95-100 90-94 85-89 80-84 75-79 50-74

Robert Parker (50,100)

“Great” “Outstanding” “Very good” “Good” “Mediocre” “Not recommended”

96-100 90-95 80-89 70-79 60-69 50-59

“Extraordinary” “Outstanding” “Above average/very good” “Average” “Below average” “Unacceptable”

Notes: For both the Wine Spectator and Robert Parker rating systems, the table reports the quality bins corresponding to the scores of the different wines.

33

ACCEPTED MANUSCRIPT Table 4: Summary Statistics on Exports Data by Year Firms

Wines

Exports (US dollars)

Observations

2002 2003 2004 2005 2006 2007 2008 2009

59 73 107 120 150 148 148 151

794 933 1,171 1,517 1,731 1,860 1,804 1,833

36,504,644 50,664,899 69,640,144 91,261,787 112,540,681 131,147,970 125,851,505 108,177,602

2,067 3,056 3,923 5,330 6,793 7,407 6,865 6,135

2002–2009

209

6,720

SC

RI P

T

Year

725,789,235

41,576

NU

Notes: For each year in the sample, the table reports the number of exporters, the number of wines exported, the value

MA

of FOB exports (in US dollars), and the number of observations.

Table 5: Summary Statistics

Maximum

Mean

Median

Std. dev.

41,576 41,576 41,576 41,576 26,705

0.02 1 1 55 72

381 510 65 97 98

5.3 139 24 85 87

3.6 120 21 85 87

6.7 130 16 3.8 2.3

ED

Minimum

PT

Unit values (US dollars per liter) Number of wines exported Number of export destinations Wine Spectator Parker

Observations

CE

Notes: For each variable listed in the first column, the table reports its minimum, maximum, mean, median value,

AC

standard deviation, and the number of observations available.

Table 6: Top Export Destinations 2002–2009 Export destinations

Share of FOB exports (%)

United States Netherlands United Kingdom Brazil Canada Denmark Finland Sweden Switzerland Germany France

30.8 10.3 9.3 7.2 6.4 6.0 3.2 3.2 2.8 2.3 1.8

Notes: For the period 2002–2009, the table reports total wine exports to each destination country as a share of total wine exports.

34

ACCEPTED MANUSCRIPT

Table 7: Snapshot of the Data Spectator 2006 United States 2006 United States 2006 China United Kingdom United Kingdom Thailand

Name

Type

Grape

Vintage year

Quality

Unit value

A B B

Red Red Red

Malbec Mezcla Mezcla

2005 2003 2003

73 92 92

3.27 13.38 11.09

C D D

Red Red Red

Cabernet Sauvignon Mezcla Mezcla

2007 2004 2004

85 98 98

4.00 43.27 24.65

T

Wine 1 1 1

Parker 2 2008 2 2008 2 2008

Destination

RI P

Year

SC

Firm

NU

Notes: The firm and wine names are confidential and are replaced by numbers and letters instead. The table reports, for individual wines (defined according to their name, grape, type, and vintage year) exported to specific destinations in specific years, their quality rating from either the Wine Spectator or Parker and their export unit value (in US dollars

MA

per liter).

Table 8: Descriptive Statistics by Quality Bin of the Wine Spectator

PT

80 5,667 16,382 16,961 2,282 204

CE

“Great” “Outstanding” “Very good” “Good” “Mediocre” “Not recommended”

Firms

ED

Observations

6 68 128 155 80 24

Export share

Unit value

Destinations per wine

Distance

GDP per capita

0.05 14.81 43.73 36.10 4.32 1.00

27.4 11.9 4.4 4.0 3.8 3.9

3 11 14 11 6 11

10,137 9,570 9,823 9,880 9,655 9,908

30,386 26,668 26,278 26,414 28,012 25,672

AC

Notes: For each quality bin of the Wine Spectator, the table reports the number of observations, the number of exporters, exports as a share of total exports (%), the mean unit value (in US dollars per liter), the number of destinations (per wine exported), the mean distance shipped (in kilometers), and the mean income per capita of the destination countries (in US dollars).

Table 9: Descriptive Statistics for Exporters by Quality Bin of the Wine Spectator

“Great” “Outstanding” “Very good” “Good” “Mediocre” “Not recommended”

Observations

Firms

Export share

Minimum quality

Maximum quality

Mean quality

Wines

Destinations per wine

3,719 27,919 8,614 1,232 91 1

6 62 80 48 12 1

16.53 67.71 13.72 2.00 0.04 0.00

77 60 55 65 72 72

97 94 89 84 79 72

89 85 83 81 77 72

180 182 74 30 6 1

16 13 8 5 3 1

Notes: The table reports, for firms classified by Wine Spectator quality bin based on the highest quality wine the firms exported, the number of observations, the number of exporters, exports as a share of total exports (%), the minimum, maximum, and mean quality exported, the number of wines, and the number of destinations (per wine exported).

35

ACCEPTED MANUSCRIPT

Table 10: Baseline Results (1)

(2)

(3)

(4)

0.059

−0.001

–

–

Panel A: Dependent variable is unit value (0.041)

0.032 a

–

ln RER × quality

–

–

R-squared

0.629

0.842

(0.003)

Panel B: Dependent variable is export volume 1.378 a

ln RER quality

0.002 a

0.002 c

(0.000)

(0.001)

0.842

0.844

1.289 b

0.868

(0.115)

3.421 a

(0.516)

(0.611)

(0.616)

(0.813)

−0.050a

–

–

–

–

−0.005b

−0.017a

0.425

0.420

1.350 c

ln RER × quality

–

ln REER

0.724

NU

(0.006)

2.167

(0.002)

(0.633)

a

(0.265)

0.567

Quality

WS

Observations

41,576

2.315

ED

R-squared

MA

(0.536)

ln GDP per capita

(0.062)

SC

quality

T

0.190 a

(0.038)

RI P

0.140 a

ln RER

(0.632)

a

2.294

(0.005)

(0.765) a

2.114 a

(0.335)

(0.334)

(0.420)

0.753

0.753

0.749

WS

WS

Parker

41,576

41,576

26,705

CE

PT

Notes: Firm-destination, firm-year, grape, type, vintage year, province, and HS fixed effects are included in (1). Firmdestination and product-year fixed effects are included in (2) to (4). Robust standard errors adjusted for clustering at the product-level are reported in parentheses. a , b , and c indicate significance at the one, five, and ten percent levels, respectively. Unit values are in pesos per liter and export volumes are in liters.

Table 11: Extensive Margin

(1)

(2)

(3)

(4)

ln RER

AC

Dependent variable is export volume 1.378 a

1.864 a

(0.516)

(0.519)

−0.050a

−0.051a

ln RER × quality

–

ln REER

0.724

quality

ln GDP per capita

(0.006)

1.986 a

(0.005)

0.012 a

2.699 a

(0.007)

0.009 a

(0.001)

(0.001)

−0.006a

–

−0.008a

0.720

2.819 a

2.831 a

(0.006) (0.002)

(0.536)

(0.535)

(0.100)

2.167 a

2.142 a

6.339 a

(0.001)

(0.100)

6.335 a

(0.265)

(0.265)

(0.006)

(0.006)

0.567

0.567

0.130

0.130

Estimator

OLS

OLS

Tobit

Tobit

Observations

41,576

41,576

998,350

998,350

R-squared/Pseudo R-squared

Notes: Firm-destination, firm-year, grape, type, vintage year, province, and HS fixed effects are included. Robust standard errors adjusted for clustering at the product-level are reported in parentheses. a , b , and c indicate significance at the one, five, and ten percent levels, respectively. Export volumes are in liters.

36

ACCEPTED MANUSCRIPT Table 12: Local Distribution Costs (1)

(2)

(3)

(4)

Panel A: Dependent variable is unit value 0.111 a

(0.057)

(0.083)

–

ln RER × quality

0.107 c

−0.155c

(0.027)

0.004 a

T

ln RER

0.001

(0.001)

0.084 a

–

–

0.017 a

–

ln dist. cost × quality

–

–

ln RER × quality × ln dist. cost

–

–

R-squared

0.936

0.965

ln RER × ln dist. cost

(0.004)

Panel B: Dependent variable is export volume ln RER

2.757 a

0.779

(0.519)

ln RER × quality

–

ln dist. cost

0.686 a

−0.025

ln dist. cost × quality

–

ln RER × quality × ln dist. cost

–

0.345

PT

(0.538)

ln GDP per capita

2.198

Observations

CE

R-squared Distribution costs

ED

(0.019)

a

0.018

(0.105)

–

−0.111b

–

0.001

–

(0.051)

(0.001)

0.001 b

(0.001)

0.920

0.937 1.270 b

0.783

(0.797)

(0.534)

−0.009b

−0.002

−0.006a

–

–

0.245

–

–

–

–

–

–

MA

(0.090)

0.001

(0.001)

(0.943)

(0.004)

ln RER × ln dist. cost

ln REER

SC

(0.017)

NU

ln dist. cost

RI P

(0.001)

0.064

(0.041)

0.645

0.733 a

(0.224)

0.005

(0.005)

−0.009a (0.003)

0.361

(0.809) a

2.828

(0.002)

(0.459)

0.378

(0.845)

2.246

(0.002)

(0.539) a

2.185 a

(0.289)

(0.485)

(0.420)

(0.290)

0.825

0.847

0.875

0.825

All

High

Low

All

40,397

19,723

20,674

40,397

AC

Notes: Firm-destination, product-year, and product-wine categories fixed effects are included. Robust standard errors adjusted for clustering at the product-level are reported in parentheses. a , b , and c indicate significance at the one, five, and ten percent levels, respectively. Unit values are in pesos per liter and export volumes are in liters.

37

ACCEPTED MANUSCRIPT

Table 13: Income per Capita (1)

(2)

(3)

(4)

(5)

Panel A: Dependent variable is unit value 2.139 a

1.240 a

(0.594)

(0.424)

ln RER × quality

−0.011c

−0.011c

ln GDP per capita

0.126

0.140

(0.006)

(0.006)

(0.152)

ln RER × ln GDP per capita

−0.173

ln GDP per capita × quality

−0.001

(0.152) a

−0.156

(0.060)

−0.001

(0.002)

ln RER × quality × ln GDP per capita

−0.011b

−0.013b

−0.015a

0.037

0.073

−0.068

−0.188a

−0.154a

(0.005)

−0.128

a

(0.046)

SC

(0.061) (0.001)

0.001 b

0.001 b

1.882 a

(0.498)

(0.104)

b

1.815 a

(0.557)

RI P

(0.548)

T

1.631 a

ln RER

−0.001 (0.001)

0.001 b

(0.006)

(0.156) (0.062)

0.001

(0.002)

0.002 c

(0.005) (0.108) (0.048)

0.001

(0.001)

0.002 a

(0.000)

(0.000)

(0.001)

(0.001)

ln RER × ln distance

–

−0.077b

–

–

−0.046c

ln dist. cost

–

–

0.084 a

–

0.081 a

ln RER × ln dist. cost

–

0.016 a

–

R-squared

0.842

0.842

0.937

0.845

0.936

−2.155

−1.361

−1.207

−4.802b

0.018

0.021

0.031

0.032

NU

(0.000)

(0.039)

MA

–

(0.017) (0.004)

(0.027)

(0.018)

0.017 a

(0.004)

Panel B: Dependent variable is export volume ln RER

−0.047

(2.418)

ED

(2.092)

ln RER × quality

0.019

ln GDP per capita

(2.104)

(0.021)

(0.021)

(0.021)

(0.022)

3.843 a

3.791 a

3.171 a

4.530 a

(0.647)

(0.648)

0.111

PT

ln RER × ln GDP per capita

(1.993)

0.044

(0.582)

0.253

(0.663)

0.231

(2.280)

(0.021)

3.695 a

(0.598)

0.290

(0.216)

(0.218)

(0.208)

(0.218)

(0.214)

−0.020a

−0.020a

−0.011c

−0.028a

−0.017a

ln RER × quality × ln GDP per capita

−0.002

−0.002

−0.003

−0.004

−0.004

ln REER

0.400

0.275

0.347

0.238

0.214

AC

ln RER × ln distance

CE

ln GDP per capita × quality

(0.007)

(0.007)

(0.002)

(0.002)

(0.006) (0.002)

(0.007) (0.002)

(0.631)

(0.643)

(0.536)

(0.645)

–

0.304 b

–

–

0.684 a

–

(0.161)

(0.006) (0.002)

(0.556)

0.354 b

(0.142)

–

–

ln RER × ln dist. cost

–

–

−0.017

–

−0.039b

R-squared

0.753

0.753

0.825

0.755

0.826

US included

Yes

Yes

Yes

No

No

Observations

41,576

41,576

40,397

36,714

35,535

ln dist. cost

(0.090)

(0.019)

0.566 a

(0.095)

(0.019)

Notes: Firm-destination and product-year fixed effects are included. Product-wine categories fixed effects are further included in (3) and (5). Robust standard errors adjusted for clustering at the product-level are reported in parentheses. a b , , and c indicate significance at the one, five, and ten percent levels, respectively. Unit values are in pesos per liter and export volumes are in liters.

38

ACCEPTED MANUSCRIPT Table 14: High versus Low Income Countries (1)

(2)

(3)

(4)

(5)

Panel A: Dependent variable is unit value 0.891 a

0.010

ln RER × quality × Low

(0.068)

(0.342)

−0.001

−0.001

a

a

(0.001)

0.002

ln RER × quality × High

(0.001)

0.002

(0.001)

(0.001)

ln RER × ln distance

–

−0.102a

ln dist. cost

–

–

ln RER × ln dist. cost

–

–

R-squared

0.842 1.003

ln RER × High

ln GDP per capita

US included Observations

0.081 a

0.015 a

–

0.937

0.845

0.936

0.001

0.001

(1.660)

(0.017)

(0.004)

1.090 c

(0.027)

(0.018)

0.017 a

(0.004)

(0.654)

(0.704)

0.988

−2.065

1.167 b

(0.537)

1.438 b

(0.650)

−2.296

0.001

−0.001

0.001

(1.366) (1.473)

(0.003)

(0.003)

(0.003)

−0.006a

−0.006a

−0.005a

−0.008a

−0.006a

0.416

0.283

0.303

0.259

0.126

(0.002)

(0.002)

(0.003) (0.002)

(0.635)

(0.654)

(0.540)

(0.650)

2.283 a

2.305 a

2.181 a

2.297 a

PT CE

R-squared

AC

ln RER × ln dist. cost

0.001 a

–

−0.849

–

ln dist. cost

0.002

(0.001)

a

0.084 a

1.384 b

(0.337)

ln RER × ln distance

−0.002

−0.042

(1.571)

(0.002)

ln REER

0.001

(0.001)

–

−0.971

ED

ln RER × quality × High

0.001

b

(0.235)

–

0.842

(0.628)

ln RER × quality × Low

(0.001)

0.378

(0.073)

(0.000)

MA

(0.722)

−0.002

0.052

0.695 a

(0.259)

(0.001)

NU

ln RER × Low

0.021

(0.040)

0.317 b

(0.124)

(0.000)

(0.038)

Panel B: Dependent variable is export volume

0.389 a

(0.145)

SC

ln RER × High

(0.330)

T

1.083 a

(0.124)

RI P

0.303 b

ln RER × Low

(0.337)

(0.291)

(0.345)

0.243

–

–

0.681 a

–

(0.162)

(0.003) (0.002)

(0.565)

2.316 a

(0.295)

0.396 a

(0.146)

–

–

–

–

−0.017

–

−0.039b

0.753

0.753

0.825

0.755

0.826

Yes

Yes

Yes

No

No

41,576

41,576

40,397

36,714

35,535

(0.090)

(0.019)

0.562 a

(0.095)

(0.019)

Notes: Firm-destination and product-year fixed effects are included. Product-wine categories fixed effects are further included in (3) and (5). Robust standard errors adjusted for clustering at the product-level are reported in parentheses. a b , , and c indicate significance at the one, five, and ten percent levels, respectively. Unit values are in pesos per liter and export volumes are in liters.

39

ACCEPTED MANUSCRIPT Table 15: Endogeneity (1)

(2)

(3)

(4)

−1.048b

–

–

Panel A: Dependent variable is unit value 0.111 c

(0.066)

(0.491)

R-squared

0.015 a

(0.000)

(0.006)

0.846

0.838

0.982

–

–

−0.004c

−0.002

−0.021b

ln REER

0.287

0.333

–

–

2.795 a

–

–

0.758

0.833

0.832

IV

OLS

IV

37,723

37,723

37,723

(2.046)

0.022

SC

ln RER × quality

(0.024)

(0.002)

2.629 a

(0.344)

R-squared

0.761 OLS 37,723

(0.333)

MA

Estimator Observations

(0.611)

NU

(0.643)

ln GDP per capita

(0.003)

0.898

−1.085

(0.632)

0.010 a

(0.001)

0.900

Panel B: Dependent variable is export volume ln RER

0.002 a

T

0.001 b

ln RER × quality

RI P

ln RER

(0.002)

(0.010)

Dependent variable

AC

ln RER

CE

Table 16: Nonlinearities

PT

ED

Notes: Firm-destination and product-year fixed effects are included. Firm-destination-year fixed effects are further included in (3) and (4). Robust standard errors adjusted for clustering at the product-level are reported in parentheses. a b , , and c indicate significance at the one, five, and ten percent levels, respectively. Unit values are in pesos per liter and export volumes are in liters. The instruments in (2) and (4) include the monthly average temperature and total rainfall per province over the growing period (September to March), as well as the altitude of each province, interacted with the exchange rate.

ln RER × quality (50 − 74) ln RER × quality (75 − 79) ln RER × quality (80 − 84) ln RER × quality (85 − 89) ln RER × quality (90 − 94)

(1)

(2)

(3)

Unit value

Unit value

Export volume

0.067

0.042

1.230

(4) c

1.290 b

(0.101)

(0.062)

0.002

–

−0.004

–

–

−0.004

–

–

−0.004

–

–

−0.004

–

–

−0.005

–

0.001 a

–

−0.005b

0.006 c

−0.002

−0.003

(0.001)

0.002 c

(0.001)

0.002

(0.001)

0.002

(0.001)

0.002 c

(0.001)

ln RER × quality (50 − 94)

–

ln RER × quality (95 − 100)

0.006 c

(0.000)

(0.003)

(0.003)

ln REER

–

–

ln GDP per capita

–

–

R-squared

0.842

Observations

41,576

(0.694)

Export volume

(0.005) (0.005) (0.004) (0.004) (0.004)

(0.004)

0.424

(0.631)

2.308 a

(0.616)

(0.002)

(0.003)

0.417

(0.632)

2.293 a

(0.334)

(0.334)

0.842

0.753

0.753

41,576

41,576

41,576

Notes: Firm-destination and product-year fixed effects are included. Robust standard errors adjusted for clustering at the product-level are reported in parentheses. a , b , and c indicate significance at the one, five, and ten percent levels, respectively. Unit values are in pesos per liter and export volumes are in liters.

40

ACCEPTED MANUSCRIPT Table 17: Productivity and the Quality of Exporters (1)

(2)

(3)

(4)

0.063

−0.002

Panel A: Dependent variable is unit value −0.013

(0.116)

0.002 a

ln RER × quality R-squared

(0.102)

0.003 a

T

(0.072)

0.001

(0.001)

(0.001)

0.818

0.870

(0.001)

RI P

ln RER

0.844

Panel B: Dependent variable is export volume 2.921 b

0.208

(0.718)

(1.282)

ln RER × quality

−0.007a

−0.002

ln REER

−0.830

2.181 c

ln GDP per capita

2.674 a

(0.002)

(0.003)

(1.262)

(0.769)

NU

1.494 a

1.267

−0.001 (0.001)

0.830 1.432 c

(0.898)

(0.851)

−0.005c

−0.006b

0.160

0.755

SC

ln RER

0.132 c

(0.075)

(0.003)

(0.916)

2.498 a

(0.003)

(0.878)

2.079 a

(0.559)

(0.481)

(0.465)

R-squared

0.656

0.823

0.763

0.742

Exporters’ quality

All

All

High

Low

Exporters’ size

Large

Small

All

All

Observations

20,819

20,757

20,721

20,855

MA

(0.414)

ED

Notes: Firm-destination and product-year fixed effects are included. Robust standard errors adjusted for clustering at the product-level are reported in parentheses. a , b , and c indicate significance at the one, five, and ten percent levels, respectively. Unit values are in pesos per liter and export volumes are in liters.

PT

Table 18: Export Market Shares (1)

(2)

(3)

(4)

Panel A: Dependent variable is unit value 0.082

(0.072)

ln RER × quality

R-squared

AC

ln RER × market share market share

CE

ln RER

0.002 a

0.009

0.184 a

(0.141)

(0.041)

0.002

–

(0.001)

(0.001)

–

–

–

–

0.888

0.857

0.009 c

(0.005)

0.114 a

0.046

(0.062)

0.002 a

(0.001)

0.010 c

(0.005)

0.115 a

(0.014)

(0.014)

0.843

0.843

Panel B: Dependent variable is export volume ln RER

1.070

1.658

0.915

1.281 b

(0.755)

(1.608)

(0.594)

(0.597)

ln RER × quality

−0.005b

−0.001

–

−0.004b

ln RER × market share

–

–

0.068 a

0.066 a

market share

–

–

ln REER ln GDP per capita

(0.002)

0.230

(0.772)

1.518 a

(0.005)

(0.016)

1.716 a

(0.057)

1.406

(1.593)

4.341 a

0.583

(0.615)

2.619 a

(0.002)

(0.016)

1.715 a

(0.057)

0.580

(0.614)

2.601 a

(0.365)

(0.736)

(0.318)

(0.318)

R-squared

0.803

0.804

0.771

0.771

Market shares

High

Low

All

All

Observations

23,619

17,957

41,576

41,576

Notes: Firm-destination and product-year fixed effects are included. Robust standard errors adjusted for clustering at the product-level are reported in parentheses. a , b , and c indicate significance at the one, five, and ten percent levels, respectively. Unit values are in pesos per liter and export volumes are in liters.

41

ACCEPTED MANUSCRIPT

T

Figure 1: Argentina’s Total Wine Exports (million US dollars) Sources: Nosis and United Nations Comtrade

AC

CE

PT

ED

MA

NU

SC

RI P

Figure 2: Comparison between the Wine Spectator and Parker ratings Figure 3: Argentinean peso per US dollar, January 2002 to December 2009 Source: International Financial Statistics

42

NU

SC

RI P

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

Figure 1: Argentina’s Total Wine Exports (million US dollars) Sources: Nosis and United Nations Comtrade

Figure 2: Comparison between the Wine Spectator and Parker ratings

42

NU

SC

RI P

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

Figure 3: Argentinean peso per US dollar, January 2002 to December 2009 Source: International Financial Statistics

43

ACCEPTED MANUSCRIPT

Highlights

CE

PT

ED

MA

NU

SC

RI P

T

Study the effects of exchange rate changes on firms heterogeneous in product quality Our model predicts more pricing-to-market for higher quality goods It also predicts that higher quality exports respond less to a real depreciation We use a data set of firm-level Argentinean wine exports to support our predictions Quality is measured using experts wine ratings

AC

x x x x x