Quality, Trade, and Exchange Rate Pass-Through
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund
  • | 2 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

Contributor Notes

Author’s E-Mail Address: n.a.chen@warwick.ac.uk and ljuvenal@imf.org

This paper investigates theoretically and empirically the heterogeneous response of exporters to real exchange rate fluctuations due to product quality. Our model shows that the elasticity of demand perceived by exporters decreases with a real depreciation and with quality, leading to more pricing-to-market and to a smaller response of export volumes to a real depreciation for higher quality goods. We test the proposed theory using a highly disaggregated Argentinean firm-level wine export dataset between 2002 and 2009 combined with experts wine rankings as a measure of quality. The model predictions find strong support in the data and the results are robust to different measures of quality, samples, specifications, and to the potential endogeneity of quality.

Abstract

This paper investigates theoretically and empirically the heterogeneous response of exporters to real exchange rate fluctuations due to product quality. Our model shows that the elasticity of demand perceived by exporters decreases with a real depreciation and with quality, leading to more pricing-to-market and to a smaller response of export volumes to a real depreciation for higher quality goods. We test the proposed theory using a highly disaggregated Argentinean firm-level wine export dataset between 2002 and 2009 combined with experts wine rankings as a measure of quality. The model predictions find strong support in the data and the results are robust to different measures of quality, samples, specifications, and to the potential endogeneity of quality.

I. Introduction

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 (or, in other words, import prices do not fully adjust to exchange rate changes).2,3 A challenge for both economists and policymakers is to understand the reasons for incomplete pass-through as the latter has implications for the implementation of optimal monetary and exchange rate policies.4 Possible explanations for partial pass-through include short run nominal rigidities combined with pricing in the currency of the destination market (Engel, 2003; 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).5

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 shocks.6 However, evidence on the role of product-level characteristics in explaining heterogeneous pass-through remains scarce. In order to fill this gap, this paper explores how incomplete pass-through can be explained by the quality of the goods exported. We model theoretically the effects of real exchange rate shocks 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. 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 expert wine ratings as a directly observable measure of quality.7

The first contribution of the paper is to develop a theoretical model to guide our empirical specifications. Building on Berman, Martín, and Mayer (2012) and Chatterjee, Dix-Carneiro, and Vichyanond (2013), we extend the model of Corsetti and Dedola (2005) by allowing firms to produce and export multiple products with heterogeneous levels of quality. In the presence of additive local distribution costs paid in the currency of the importing country, the model shows that the demand elasticity perceived by the firm falls with a real depreciation and with quality. As a result, following a change in the real exchange rate, exporters change their prices (in domestic currency) more, and their export volumes less, for higher quality products. Once we allow for higher income countries to have a stronger preference for higher quality goods, as the evidence from the empirical trade literature tends to suggest (Crinò and Epifani, 2012; Hallak, 2006), the heterogeneous response of prices and quantities to exchange rate changes due to quality is predicted to be stronger for higher income destination countries.

The second contribution of the paper is to bring the predictions of the model to the data. The firm-level trade data we rely on are from the Argentinean customs which provide, for each export flow between 2002 and 2009, the name of the exporting firm, the country of destination, the date of the shipment, the Free on Board (FOB) value of exports (in US dollars), and the volume (in liters) of wine exported. The level of disaggregation of the data is unique because for each wine we have its name, grape (Chardonnay, Malbec, etc.), type (white, red, or rosé), and vintage year. With such detailed information we can define a “product” in a much more precise way compared to the papers that rely on trade classifications such as the Harmonized System (HS) to identify products. For instance, Argentina’s 12-digit HS classification only groups wines into eleven different categories or “products.” In contrast, as we define a product according to the name of the wine, its grape, type, and vintage year, the sample we use for the estimations includes 6,720 different wines exported by 209 wine producers. The exporters in the sample are therefore multi-product firms.

In order to assess the quality of wines we rely on two well-known experts wine ratings, the Wine Spectator and Robert Parker. In both cases a quality score is assigned to a wine according to its name, grape, type, and vintage year which are characteristics we all observe in the customs data so the trade and quality data sets can directly be merged with each other. Quality is ranked on a (50,100) scale with a larger value indicating a higher quality. Our approach to measuring quality is similar to Crozet, Head, and Mayer (2012) who match French firm-level export data of Champagne with experts quality assessments to investigate the relationship between quality and trade. However, in contrast to our paper they are unable to distinguish between the different varieties sold by each firm, so each firm is assumed to export one type of Champagne only.

We compute FOB export unit values as a proxy for export prices at the firm-product-destination level, and investigate the pricing strategies of exporters in response to real exchange rate fluctuations between trading partners (i.e., between Argentina and each destination country). Consistent with other firm-level studies, we find that pass-through is large: in our baseline regression, following a ten percent change in the real exchange rate exporters change their export prices (in domestic currency) by 1.1 percent so pass-through is 89 percent. Also, as expected, we find that higher quality is associated with higher prices. Most interestingly, we show that the response of export prices to real exchange rate changes increases with the quality of the wines exported, or in other words pass-through decreases with quality. A one standard deviation increase in quality from its mean level increases pricing-to-market by five percent. Also, pass-through is complete (i.e., 100 percent) for the wine with the lowest quality in the sample, but drops to 86.5 percent for the highest quality wine. This heterogeneity in the response of export prices to exchange rate changes remains robust to different measures of quality, samples, and specifications. We also examine the heterogeneous response of export volumes to real exchange rate fluctuations. Export volumes increase following a real depreciation, but by less for higher quality goods. Finally, we find that the response of export prices (volumes) to real exchange rate changes increases (decreases) with quality only when firms export to high income destination countries. Overall, our empirical results find strong support for the predictions of the model.

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 and this might affect in a way or the other the scores they assign to the different wines, leading to an endogeneity bias. In order to overcome this issue, we use appropriate instruments for quality based on geography and weather-related factors, including the total amount of rainfall and the average temperatures during the growing season for each province where the grapes are grown, as well as the altitude of each of the growing regions of Argentina. We show that our main findings remain robust to the instrumentation of quality.

The degree of exchange rate pass-through of 89 percent that we find, which magnitude is consistent with the estimates of other firm-level studies, contrasts with the low pass-through that is typically estimated using aggregate or industry-level data. For instance, in a sample of OECD countries, Campa and Goldberg (2005) find an average pass-through of 46 percent in the short run and 64 percent in the long run. We therefore investigate whether pass-through estimates suffer from an aggregation bias. To this aim, we aggregate our data both at the firm and at the country-levels, re-estimate our benchmark specifications, and compare the magnitude of exchange rate pass-through estimated at each level of data aggregation. Interestingly, we find that the more aggregated the data, the lower is estimated pass-through, suggesting that aggregate pass-through estimates suffer from an aggregation bias.

Our paper belongs to two strands of the literature. The first one is the vast literature on incomplete exchange rate pass-through and pricing-to-market. Among the papers that explore the determinants of heterogeneous pass-through from the perspective of exporting firms, Berman and others (2012) find that highly productive French exporters change significantly more their export prices in response to real exchange rate changes, leading to lower pass-through. Chatterjee and others (2013) focus on multi-product Brazilian exporters and show that within firms, pricing-to-market is stronger for the products the firm is most efficient at producing. Amiti, Itskhoki, and Konings (2012) find that Belgian exporters with high import shares and high export market shares have a lower exchange rate pass-through.8

Our paper is also related to Auer and Chaney (2009) and Auer, Chaney, and Sauré (2012) who explore the relationship between quality and pass-through. However, as the two papers rely on import and consumer prices data, respectively, their empirical analysis investigates exchange rate pass-through rather than the pricing-to-market behavior of exporting firms. Consistent with our paper, these authors predict that pass-through should be higher for lower quality goods.9 Auer and Chaney (2009) do not find any evidence for such a relationship using import prices data for the US, where quality is inferred from trade unit values. In contrast, using a data set on the prices and numbers of cars traded in Europe, Auer and others (2012) find some evidence that pass-through decreases with hedonic quality indices estimated from regressions of car prices on car characteristics such as weight, horse power, and fuel efficiency.

Second, this paper relates to the growing literature on quality and trade, which mostly relies on trade unit values in order to measure quality.10 At the country level, Hummels and Klenow (2005) and Schott (2004) focus on the supply-side and show that export unit values are increasing in exporter per capita income. On the demand-side, Hallak (2006) finds that richer countries have a relatively stronger demand for high unit value exporting countries. More recently, some papers have started to 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 some 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. Closest to our work is Crozet and others (2012) who explain French firm-level export prices and quantities of Champagne by experts ratings as a measure of quality.11

The paper is organized as follows. In section 2 we present our model where firms export multiple products with heterogeneous levels of quality, and show how real exchange rate changes affect the optimal price and quantity responses of exporters. Section 3 describes our firm-level exports customs data, the wine experts quality ratings, and the macroeconomic data we use. Section 4 discusses how the features of the wine industry conform with the main assumptions of the theoretical model to be tested. Section 5 presents our main empirical results. Section 6 provides robustness checks, while section 7 concludes.

II. A Model of Pricing-to-Market and Quality

Berman and others (2012) extend the model with distribution costs of Corsetti and Dedola (2005), allowing for firm heterogeneity where single-product firms differ in their productivity. They show that the elasticity of demand perceived by the exporter falls with a real depreciation and productivity, leading to variable markups which increase with a real depreciation and productivity. This leads to heterogeneous pricing-to-market where more productive exporters change their prices more than others following a change in the real exchange rate.12 In their appendix, Berman and others (2012) show that a similar result holds if firms differ in the quality of the (single) good they export: firms that export higher quality goods change their export prices more than others in response to a real exchange rate change.

Chatterjee and others (2013) extend the model of Berman and others (2012) to multi-product firms. Inspired by Mayer, Melitz, and Ottaviano (2011), each firm is assumed to be most efficient at producing a key variety which is the firm’s “core competency,” and the further away a variety is from the core, the relatively less efficient each firm is at producing this variety.13 In response to a change in the real exchange rate, exporters vary their prices more for the products closer to their core competency, which in turn have a higher efficiency and therefore smaller marginal costs.

In what follows, we build on Berman and others (2012) and Chatterjee and others (2013) and extend the model of Corsetti and Dedola (2005), allowing for firm heterogeneity in the quality of the goods exported. Given that most firms in our data set export multiple products, we model them as multi-product firms which therefore differentiates us from Berman and others (2012) who focus on single-product firms. In contrast to the multi-product firms model of Chatterjee and others (2013), we however rank the different goods produced by each firm in terms of quality rather than efficiency, where higher quality is associated with higher marginal costs (Crinò and Epifani, 2012; Hallak and Sivadasan, 2011; Johnson, 2012; Kugler and Verhoogen, 2012; Manova and Zhang, 2012a; Verhoogen, 2008). We then look at how changes in real exchange rates affect the optimal price and quantity responses of exporters and derive some testable implications that can be taken to the data.

A. The Basic Framework

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 by14

U(Cj)=[Ψ[s(φ)xj(φ)]σ1σdφ]σσ1,(1)

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. The set of available varieties is 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.

Firms are multi-product and produce goods with different levels of quality. They are heterogeneous in two dimensions: efficiency/productivity and product quality. The parameter φ, which denotes each variety, indicates how efficient each firm is at producing each variety so φ has both a firm- and a product-specific component. Each firm produces one “core” product, but in contrast to Chatterjee and others (2013) or Mayer and others (2011) who consider that a firm’s core competency lies in the product it is most efficient at producing – and which therefore has lower marginal costs – we assume that a firm’s core competency is in its product of superior quality which entails higher marginal costs (Manova and Zhang, 2012b).

The efficiency associated with the core product is given by a random draw Φ so 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 then observe a hierarchy of products based on their quality levels. A firm with core efficiency Φ then produces a product r with an efficiency level φ given by

φ(Φ,r)=Φϑr,(2)

where ϑ > 1. Products with smaller r (higher quality) are closer to the core and therefore have a lower efficiency φ (Φ, r). Higher quality goods have a lower efficiency because they have higher marginal costs

s(φ(Φ,r))=(wφ(Φ,r))λ,(3)

where λ > 1 implies that markups increase with quality and w is the wage of the Home country (Berman and others, 2012).15 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. The latter captures wholesale and retail costs to be 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 ηjWjs (φ (Φ, r)). As in Berman and others (2012), we assume that higher quality goods have higher 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 and by the efficiency of the exporter in producing each good.

In units of currency of country j, the consumer price in j of a variety exported from Home to j is

pjc(φ)pj(φ(Φ,r))τjɛj+ηjwjs(φ(Φ,r)),(4)

where pj(φ) 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 will lead to a less than proportional change in the consumer price pjc(φ) (i.e., incomplete pass-through) given that local distribution costs are unaffected by currency fluctuations.16 The quantity demanded for this variety in country j is

xj(φ)=YjPjσ1[pj(φ(Φ,r)τjs(φ(Φ,r))ɛj+ηjwj]σ,(5)

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

cj(φ)=wxj(φ(Φ,r))τjφ(Φ,r)+Fj.(6)

Expressed in Home currency, the profit maximizing export price for each product the firm exports to country j is

pj(φ)=σσ1(1+njqjφ(Φ,r)s(φ(Φ,r))στj)wφ(Φ,r)=m(φ(Φ,r))wφ(Φ,r),(7)

where qjɛjwj/w is the real exchange rate between Home and j. In contrast to the standard Dixit-Stiglitz markup (Dixit and Stiglitz, 1977), the presence of local distribution costs leads to variable markups m (φ(Φ, r)) over marginal costs that are larger than σσ1, increase with quality s (φ(Φ, r)), the real exchange rate qj (i.e., a real depreciation), and local distribution costs ηj.18

The volume of exports xj (φ) is given by

xj(φ)=(σ1σ)σYjPjσ1[wφ(Φ,r)s(φ(Φ,r))ɛjτj+ηjwj]σ(8)

so the elasticity (in absolute value) of the exporter’s demand xj(φ) with respect to the export price pj (φ) is

ej=|xj(φ)pj(φ)pj(φ)xj(φ)|=στj+ηjqjφ(Φ,r)s(φ(Φ,r))τj+ηjqjφ(Φ,r)s(φ(Φ,r)),(9)

which is decreasing in quality and with a real depreciation. For a product that is closer to the core, quality is higher, the elasticity of demand is smaller, and the markup is higher. The model leads to two predictions on the effects of exchange rate changes on export prices and quantities that can be tested in the data.

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 and which captures the degree of pricing-to-market, increases with the quality of the good exported, s(φ(Φ, r)):

epj=|pj(φ)qjqjpj(φ)|=ηjqjφ(Φ,r)s(φ(Φ,r))στj+ηjqjφ(Φ,r)s(φ(Φ,r)).

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 good exported, s(φ(Φ, r)):

exj=|xj(φ)qjqjxj(φ)|=στjτj+ηjqjφ(Φ,r)s(φ(Φ,r)).

Intuitively, the mechanism is the following. A real depreciation reduces the elasticity of demand perceived by exporters in the destination country, which allows all firms to increase their markups. As higher quality goods have a smaller elasticity of demand, their markups can therefore be increased by more than for lower quality goods. This leads to heterogeneous pricing-to-market which is stronger for higher quality goods (i.e., pass-through is lower). In turn, this implies that the response of export volumes to a real depreciation decreases with quality. This mechanism is similar to Berman and others (2012) and Chatterjee and others (2013), although their focus is on productivity differences in driving heterogeneous pricing-to-market across exporters, or exporters and products, respectively.

B. Cross-Country Heterogeneity in the Preference for Quality

In the previous section, we assumed that the preference for quality is homogeneous across destination countries. The evidence in the literature however suggests that consumer preferences for quality may vary from one country to the other as preferences are affected by per capita income. In particular, consumers in richer countries are expected to have stronger preferences for higher quality products so the consumption of higher quality goods is increasing in per capita income.19 Hallak (2006) finds that rich countries tend to import relatively more from countries that produce higher quality goods. We therefore extend the model to allow for non-homothetic preferences for quality.20

Let us assume that the Home country now exports to only two destinations f, where f is either high or low income. We build on Crinò and Epifani (2012) and assume that the preference for quality is increasing in per capita income. The utility function becomes (also, see Hallak, 2006)

U(Cf)=[Ψ[s(φ)ι(yf)xf(φ)]σ1σdφ]σσ1,(10)

where ι (yf) captures the intensity of preference for quality with respect to per capita income yf, and yhigh > ylow so countries with higher per capita income have a stronger preference for quality. Local distribution costs are thus higher in high income countries as ηfwfs(φ)ι(yf) increases in per capita income.21,22 This allows us to derive two additional predictions that can be tested in the data.

Prediction 3. The firm- and product-specific elasticity of the export price pf (φ) to a change in the real exchange rate qf, denoted by epf, increases with the quality of the good exported s(φ(Φ, r)), and by more for high income than for low income destination countries:

epf=|pf(φ)qfqfpf(φ)|=ηfqfφ(Φ,r)s(φ(Φ,r))ι(yf)στf+ηfqfφ(Φ,r)s(φ(Φ,r))ι(yf).

Prediction 4. The firm- and product-specific elasticity of the volume of exports xf (φ) to a change in the real exchange rate qf, denoted by exf, decreases with the quality of the good exported s(φ(Φ, r)), and by more for high income than for low income destination countries:

exf=|xf(φ)qfqfxf(φ)|=στfτf+ηfqfφ(Φ,r)s(φ(Φ,r))ι(yf).

III. Data and Descriptive Statistics

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

A. Firm-Level Exports Customs Data

Before the 1990s, Argentinean wines were rarely exported to international markets. Since then, wine exports started to gain strength thanks to the successful strategies implemented by one of the main wine producers, Nicolás Catena Zapata.23 Catena played a key role in making Argentinean wines internationally recognized, and the growth in the wine sector that followed was hence spectacular: by the mid-2000s, Argentina was the eighth largest wine exporter and the fifth wine producer in the world.24 During the 2000s, the sector continued to boom and exports more than tripled between 2002 and 2009.

The firm-level exports data we use are from the Argentinean customs and are provided to us by a private vendor called Nosis. For each export flow we have 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 2009.25 We also have the name/brand of the wine exported, its type (red, white, or rosé), grape (Malbec, Chardonnay, etc.), and vintage year.26 Figure 1 compares the total value of Argentina’s wine exports from our customs data set with the value reported in the Commodity Trade Statistics Database (Comtrade) of the United Nations (HS code 2204). The data coincide extremely well.

Figure 1.
Figure 1.

Argentina’s Total Wine Exports

(million US dollars)

Citation: IMF Working Papers 2014, 042; 10.5089/9781475526394.002.A001

Given that actual export prices are not available we proxy for them using the unit values of exports in local currency, computed as the ratio of the export value in Argentinean pesos divided by the corresponding export volume in liters.27 In order 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.

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.” Second, we only keep the export flows recorded as FOB.28 Third, as the experts rankings we rely on to measure quality are only for red, white, or rosé wines, we drop all sparkling wines, dessert wines, and other special varieties. Fourth, as we are interested in how product quality affects the pricing and export decisions of firms, and in turn need to control for the performance of wine exporters in the regressions, we restrict our analysis to wine producers and therefore to the manufacturing sector only – which requires us to drop wholesalers and retailers. The Instituto Nacional de Vinticultura’s (INV), the government’s controlling body for the wine industry, provides us with the names of all the firms authorized to produce and sell wine, as well as their activity classification. We match the exporters names from the customs data with the list of firms provided by the INV and only keep wine producers. Fifth, we drop a number of typos which we are unable to fix. For instance, we exclude the very few cases where the vintage year reported is ahead of the year in which the exports took place. We also drop the few observations where the value of exports is positive but the corresponding volume is zero. Finally, we also exclude a few outliers: for each exporter, we drop the observations where unit values are larger or smaller than 100 times the median export unit value charged by the firm.

The recent papers on heterogeneous pass-through typically define a “product” according to trade classifications such as the Harmonized System or the Combined Nomenclature (e.g., Amiti and others, 2012; Auer and Chaney, 2009; Berman and others, 2012; Chatterjee and others, 2013). As Table 1 shows, the 6-digit HS classification categorizes wines into four different categories according to whether they are sparkling or not, and to the capacity 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.29 The problem is that changes in unit values defined at this level may reflect compositional changes rather than price changes as there may be more than one distinct product within a single HS code. In contrast, the detail provided by our data set allows us to define an individual 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 changes are unlikely to affect unit values.30 Our cleaned sample includes a total of 21,647 different products/wines of which 6,720 can be matched with quality rankings. The 6,720 wines only represent 31 percent of all wines, but 58 percent of the total FOB value exported between 2002 and 2009.31

Table 1.

Harmonized System (HS) Classification Codes

article image

We close this section with descriptive statistics on the sample we use for the estimations. Table 2 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 different wine producers.32 As shown by Table 3, these firms exported an average of 139 different wines, ranging from a minimum of one to a maximum of 510 (in the sample, only 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 rankings). Exporters charged between two cents and 381 US dollars per liter of wine exported, with an average of five US dollars per liter. Firms exported to an average number of 40 different destinations, from a minimum of one to a maximum of 88. Table 4 shows that with the exception of Brazil, Argentinean wine exporters mostly sell to developed economies, the United States being the top destination market.

Table 2.

Summary Statistics on Trade Data by Year

article image
Notes: Authors’ own calculations.
Table 3.

Summary Statistics

article image
Notes: Authors’ own calculations.
Table 4.

Top Export Destinations 2002-2009

article image
Notes: Authors’ own calculations.

B. Quality Ratings

The editors of the Wine Spectator magazine review more than 15,000 wines each year in blind tastings and publish their rankings in several issues throughout the year.33 The rankings 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. Table 5 lists the six different categories the wines fall in depending on the score they are given.

Table 5.

Experts Ratings

article image

We match the wines from the customs data set with the ones reviewed by the Wine Spectator by name, type, grape, and vintage year so that each wine is assigned a single quality ranking.34 We end up with 6,720 wines exported by 209 firms over the 2002-2009 period. As can be seen from Table 3, the mean ranking is 85, the lowest-rated wine receives a score of 55, and the highest receives a score of 97. The distribution across wines is very symmetric as the mean and the median are both equal to 85. Note that our approach to measuring quality is similar to Crozet and others (2012) who match French firm-level exports data of Champagne with experts quality assessments in order to investigate the relationship between quality and trade. However, due to data limitations they are unable to distinguish between the different varieties sold by each firm so each firm is assumed to export one type of Champagne only. In addition, their ratings are only measured on a (1,5) scale, where a larger value indicates a higher quality.

We rely on the Wine Spectator for our baseline regressions because it has the largest coverage of Argentinean wines. However, in the robustness section we check the sensitivity of our results using an alternative rating produced by Robert Parker.35 Parker is a leading US wine critic who assesses wines based on blind tastings and publishes his consumer advice and rankings in a bimonthly publication, the Wine Advocate. His rating system also employs a (50,100) point scale where wines are ranked according to their name, type, grape, and vintage year, and where a larger value indicates a higher quality. Table 5 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).36 We match the customs data and the Parker rankings for 3,969 wines exported by 181 firms. Table 3 shows that the scores vary between 72 and 98 with an average of 87. Again, the distribution across wines is very symmetric as the mean and the median are equal. Figure 2 plots the Wine Spectator and Parker rankings. A total of 2,433 wines exported by 135 firms have rankings from both sources. The correlation between the two rankings is 0.53.

Figure 2.
Figure 2.

Wine Spectator versus Parker rankings

Citation: IMF Working Papers 2014, 042; 10.5089/9781475526394.002.A001

Table 6 provides a snapshot of our data. For confidentiality reasons we cannot report the exporter nor the wine names so these are replaced by numbers and letters instead. The table shows that, whether we use the Wine Spectator or the Parker ratings, individual firms export wines with varying levels of quality (between 68 and 86 for Firm 1 and 74 and 90 for Firm 2). In addition, higher quality wines are, on average, sold at a higher price. Finally, the table illustrates that the Law of One Price fails: in 2009, Firm 1 exported the same wine to two different destinations, but charged 17 US dollars per liter to Norway versus 6.6 dollars per liter to China. Similarly, in 2006 Firm 2 charged 4.4 dollars to the Netherlands versus 3.8 dollars to Brazil for the same liter of wine exported to both destinations.

Table 6.

Snapshot of the Data

article image

C. Macroeconomic Data

The data on GDPs are from the Penn World Tables, and the consumer price indices (CPI) and nominal exchange rates 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 so an increase of the exchange rate captures a real depreciation of the peso. The nominal exchange rates are available for each country relative to the US dollar, which we convert to be relative to the Argentinean peso. The real effective exchange rates are sourced from the IFS and the Bank of International Settlements where an increase indicates a real depreciation.

During the 2002-2009 period, Argentina witnessed major nominal exchange rate fluctuations. Figure 3 illustrates the evolution of the monthly nominal exchange rate between the Argentinean peso and the US dollar. After the financial crisis of 2001, the fixed exchange rate system was abandoned and as a result the peso depreciated in 2002 by up to 75 percent. The export boom that followed lead to a massive inflow of US dollars into the economy which helped to depreciate the US dollar compared to the peso. The peso then remained stable until 2008 when it depreciated again with the advent of the global financial crisis and the increase in domestic inflation.

Figure 3.
Figure 3.

Argentinean peso per US dollar, January 2002 to December 2009

Citation: IMF Working Papers 2014, 042; 10.5089/9781475526394.002.A001

IV. Wine and Model Assumptions

The model described in section 2 intends to capture a general relation between quality and pass-through which could hold for any particular market. The reason why we analyze the wine market is because we have an observable measure for quality. Although the model is general, it is instructive to see how the features of the wine industry conform with its main assumptions. First, as already discussed and illustrated by Table 6, higher quality wines tend to be exported at a higher price which is consistent with equation (7) of the model.

Second, the model assumes that higher quality wines have higher marginal costs (equation 3). Although the quality of wine depends predominantly on the quality of the grapes which is itself mostly affected by geography and weather-related factors, higher quality wines can be expected to have higher marginal costs (see Crozet and others, 2012, on Champagne). First, higher quality wines may require higher quality and therefore more expensive inputs (Johnson, 2012; Kugler and Verhoogen, 2012; Manova and Zhang, 2012a; Verhoogen, 2008). For instance, wine producers can choose more or less costly additives to be added during the winemaking process (in the various stages of fermentation or as preservatives). Second, achieving higher quality wines may depend on the production methods chosen by producers. One example is to use oak barrels for the ageing and fermentation of wine. Due to the cost of the oak and to the short lifetime of the barrels (the oak flavors of the barrels last for three or four vintages only), these barrels turn out to be very expensive and are therefore reserved to producing higher quality wines only.37 Another example is to use “drip irrigation” which allows producers to limit the yield and therefore increase the potential quality of grapes, but this system is expensive to install. Finally, there is some evidence that in order to produce higher quality wines, Argentinean wineries often produce their own grapes for their best wines (which may need to be pruned and trimmed carefully, requiring more skilled labor), and rely on suppliers for their lower quality wines (Artopoulos, Friel, and Hallak, 2011).

More direct evidence on the positive relationship between price (quality) and marginal costs can be found in Table 7 which breaks down into several components the price of non-EU wines sold in UK retail outlets (Joseph, 2012).38 The last row of the table shows that the amount that goes to the winemaker, which mainly reflects the costs of producing the wine as well as the costs of the bottle, closure, and carton, clearly increases with the price, and therefore most likely with the quality of the wine.39 We were unable to find a similar breakdown for Argentinean wines, but we believe that these figures for non-EU wines should still provide us with some useful insights on the composition of Argentinean wine prices sold in the UK.

Table 7.

Price Breakdown for Non-EU Wine sold in Retail Outlets in the UK

article image
Source: Joseph (2012).

Third, the model assumes that higher quality wines have higher distribution costs (distribution costs ηjWjs (φ(Φ, r)) increase with quality s (φ(Φ, r))). This is confirmed by the fourth row of Table 7 that shows that distribution costs amount to £0.11 for a £5.76 bottle, and increase to £0.21 for a £7.19 bottle, £0.40 for a £8.83 bottle, and to £0.51 for a £10.09 bottle.

Finally, equation (7) predicts that higher quality wines have higher markups. The second row of Table 7 shows indeed that the margin charged by the retailer increases systematically with the price of the wine (and, therefore, with quality too).40 The margin is £1.92 for a £5.76 wine and increases to £3.36 for a £10 wine. 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 £5 wine sold on the UK market, the producer markup is estimated to be approximately £0.40 and to increase to about £10 for a £25 bottle.41

We therefore conclude that the features of the wine industry closely satisfy the key assumptions of the model: higher quality wines tend to be exported at a higher price, and are characterized by higher marginal costs, distribution costs, and markups, both at the retail and producer levels.

V. Empirical Framework

Prediction 1 states that following a real depreciation, exporters increase their export price and this increase is larger the higher quality is. In order to check whether this relationship holds in the data, we estimate the following reduced-form regression

lnUVij,tk=β1lnqj,t+β2sWSk+β3lnqj,t×sWSk+β4zi,t+Ψt+μij+θgrape+ζtype+γvintage+ρHS+κp+ɛij,tk(11)

where UVij,tk is the export unit value of firm i exporting a product k to destination country j in year t, expressed in pesos per liter of wine exported and is our proxy for export prices. qj,t is the average real exchange rate between Argentina and country j in year t (an increase in qj,t captures a real depreciation). The quality of wine k is denoted by sWSk where the WS index refers to the Wine Spectator rankings. Given the level of disaggregation of the data, changes in real exchange rates are assumed to be exogenous to the pricing (and quantity) decisions of individual firms.

The export price in the exporter’s currency is a markup over marginal costs (Knetter, 1989, 1993). As a result, in order to identify a pricing-to-market behavior which requires markups to respond to exchange rate changes, the regression needs to control for firm-specific marginal costs which we denote by zi,t.42 Without any additional information on the exporters, we rely on a number of proxies that have been shown in the literature to correlate strongly with productivity/marginal costs. First is the average size of the firm, sizei,t, measured by the total volume of FOB exports by each firm in each year. Second is the total number of destination countries where each firm exports to in each year, desti,t.43 Besides, we include year fixed effects ψt to control for aggregate shocks that are common to all Argentinean exporters. We perform within estimations by including firm-destination μij fixed effects. As product fixed effects cannot be included (they are perfectly collinear with quality), we instead control for product characteristics by including grape θgrape, type ζtype, vintage year γvintage, HSρHS, and province p of origin κp fixed effects. Fixed effects for the wine names/brand are not included as they are collinear with the firm fixed effects (because each brand is sold by one firm only). β1, β2, β3, and β4 are coefficients to be estimated and ɛij,tk is an error term. Given that 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. Finally, as quality takes on a single value for each product, robust standard errors are adjusted for clustering at the product level.

Following a real depreciation, exporters are expected to increase their markups and therefore their export prices so β1 should be positive.44 Higher quality is expected to increase export prices so β2 should be positive, too. The coefficient of interest is β3, the coefficient on the interaction between the real exchange rate and quality which captures heterogeneous pricing-to-market. According to Prediction 1, the response of unit values to a real depreciation should increase with quality in which case β3 should be positive.

Prediction 2 relates to export volumes. It states that following a real depreciation, exporters increase their volume of exports but by less for higher quality products. To test this prediction we estimate

lnXij,tk=α1lnqj,t+α2sWSk+α3lnqj,t×sWSk+α4zi,t+α5Zj,t+Ψt+μij+θgrape+ζtype+γvintage+ρHS+κp+ɛij,tk(12)

where Xij,tk is the FOB export volume (in liters) of firm i exporting a product k to destination country j in year t.45 To be consistent with standard gravity models we include destination-year specific variables Zj,t such as destination country’s real GDP (deflated using each country’s CPI), GDPj,t, and real effective exchange rate Qj,t as a proxy for country j’s price index (Berman and others, 2012). If a real depreciation increases exports, α1 should be positive. If this increase is smaller for higher quality products, the coefficient on the interaction term α3 should be negative.

A. Baseline Results

Panel A of Table 8 reports the results of estimating equation (11) for unit values. Column (1) only includes the exchange rate, quality, and firm size as regressors and shows that higher quality wines are sold at a higher price, which is consistent with equation (7) and with the empirical findings of Crozet and others (2012) for Champagne.46 When the real exchange rate fluctuates, exporters significantly change their export prices: following a ten percent depreciation they raise their prices (in pesos) by 1.1 percent so that on average pass-through is 89 percent. The large degree of pass-through we find for the wine industry is therefore consistent with the findings of other papers that use firm-level data for the whole manufacturing sector. For instance, pass-through is estimated at 92 percent for French exporters (Berman and others, 2012), 94 percent for Chinese exporters (Li and others, 2012), 77 percent for Brazilian exporters (Chatterjee and others, 2013), 86 percent for Danish exporters (Fosse, 2012), and at 79 percent for Belgian exporters (Amiti and others, 2012).

Table 8.

Baseline Results

article image
Notes: Year, province, firm-destination, grape, type, vintage year, and HS fixed effects are included in (1)-(4) while year-destination and product fixed effects are included in (5). Robust standard errors are adjusted for clustering at the product level. t-statistics in parentheses. a, b, and c indicate significance at 1, 5, and 10 percent levels, respectively. Unit values are in pesos per liter and export volumes are in liters.

The estimated coefficient on the exchange rate reported in column (1) however hides a significant amount of heterogeneity in the degree of pass-through across products. To see this, column (2) adds the interaction term between the exchange rate and quality. Its estimated coefficient is positive and significant which is evidence of heterogeneous pass-through, lending support to Prediction 1 in that the elasticity of export prices to a real depreciation increases with quality. These results are consistent with the theoretical predictions of Auer and Chaney (2009) and the empirical results of Auer and others (2012).

In column (3) we use the number of export destinations for each firm as an alternative proxy for firm productivity. Its estimated coefficient is not significant, but most importantly our main conclusions regarding heterogeneous pass-through remain unaffected.47 Column (4) restricts the sample to multi-product firms, where a multi-product firm is defined as a firm-destination-year triplet with strictly more than one wine exported, and the results also remain unchanged.

In column (5) we check if our results hold in a difference-in-difference specification which includes destination-year and product fixed effects instead of the ones specified in equation (11). Both the exchange rate and quality drop from the regression, but the interaction term between the exchange rate and quality can be estimated. Although its estimated coefficient decreases in magnitude and in significance, it still indicates that the elasticity of export prices to exchange rates is larger for higher quality wines.

Panel B of Table 8 reports the results of estimating equation (12) for export volumes. From column (1) export volumes react positively to a real depreciation. The elasticity is large and equal to 1.844, which is consistent with evidence in the literature that the trade elasticities for emerging economies are generally larger than for developed countries.48 The coefficient on quality is negative and significant while the literature usually points to a positive relationship between trade and quality (for example, see Crozet and others, 2012). One crucial difference between our regressions and, for instance, Crozet and others (2012), however, is that we estimate the within-firm effect of quality on export volumes. The negative coefficient on quality therefore indicates that when a firm exports several wines with different levels of quality to a given destination, the high quality wines are on average exported in smaller quantities than the low quality wines. This is consistent with San Martín, Troncoso, and Brümmer (2008) who observe that more sophisticated, high quality wines are generally produced in smaller quantities.

The interaction between the exchange rate and quality is included in column (2). Consistent with Prediction 2, it is negative and significant suggesting that the response of export volumes to exchange rates decreases with quality. This finding remains robust to the use of the number of export destinations as a measure of firm performance (column 3) and to restricting the sample to multi-product firms (column 4). The difference-in-difference specification in column (5) does not provide evidence of an heterogeneous response of export volumes to exchange rates driven by quality.49

For each regression in Table 8 we report a quantitative evaluation of the economic effects of quality. The lower parts of Panels A and B report the change in the exchange rate elasticities following a one standard deviation increase in quality from its mean level (i.e., a four point increase on the quality scale). In column (2) of Panel A, the unit values elasticity increases from 0.115 to 0.121 which corresponds to a five percent increase in pricing-to-market. If we calculate the elasticity (not reported) for the lowest quality wine in the sample (with a ranking of 55), the elasticity is equal to 0.065 and is insignificantly different from zero, suggesting full pass-through. In contrast, for the highest quality wine (with a score of 97) the elasticity is positive and equal to 0.135 (significant at the one percent level) so pass-through drops to 86.5 percent. In column (2) of Panel B, a one standard deviation increase in quality reduces the volume elasticity by one and a half percent from 1.833 to 1.808. The elasticity for the highest quality wine is equal to 1.75 and increases to 2.02 for the lowest quality wine. Overall, the effects of quality on the price and volume elasticities remain very similar across all specifications reported in Table 8.

Recall that a key prediction of the model of Corsetti and Dedola (2005), and of our extension to their model, is that pricing-to-market increases with local distribution costs in the importing economy. In turn this implies that the difference in pass-through between high and low quality wines should increase with distribution costs. Berman and others (2012) use the data on distribution costs computed by Campa and Goldberg (2010) for 21 countries and 29 industries between 1995 and 2001, and find that the response of unit values to a real depreciation increases with local costs, especially for high productivity firms.

In order to explore this prediction of the model we also rely on Campa and Goldberg’s (2010) distribution costs data. Given these are only available between 1995 and 2001 we compute the average distribution costs over time for each of the 21 destination countries and for the “Food products and beverages” industry. Our measure for distribution costs, dcj, is therefore destination-specific, and given the limited number of countries for which the data are available the resulting sample size is reduced by half. We estimate regressions (11) and (12) and include an interaction term between the real exchange rate and (log) distribution costs. The results are reported in Table 9. For unit values, column (1) shows that the interaction between the real exchange rate and distribution costs is positive and significant at the ten percent level, suggesting that pricing-to-market increases with local costs which is consistent with the findings of Berman and others (2012) and Campa and Goldberg (2010). Column (2) further includes the interaction between the real exchange rate and quality which is positive and significant. Consistent with expectations, this indicates that the difference in pass-through between high and low quality wines increases with the size of local costs. The results for export volumes are reported in columns (3) and (4) of Table 9. The interaction between the exchange rate and local costs is negative, as expected, but is only significant at the 11 percent level.

Table 9.

Local Distribution Costs

article image
Notes: Year, province, firm-destination, grape, type, vintage year, and HS fixed effects are included. Robust standard errors are adjusted for clustering at the product level. t-statistics in parentheses. a, b, and c indicate significance at 1, 5, and 10 percent levels, respectively. Destination-specific distribution costs dcj for the “Food products and beverages” industry are from Campa and Goldberg (2010).

B. Heterogeneity across Destination Countries

Predictions 3 and 4 state that the effects described by Predictions 1 and 2 for unit values and export volumes, respectively, should be stronger for high income than for low income destination countries. This section investigates whether the two predictions can be validated by the data.

The destination countries included in our data set are split between high and low income according to the World Bank’s classification based on GNI per capita in 2011. Low income countries have a GNI per capita of less than $4,035 while high income countries are above that threshold. We then estimate equations (11) and (12) for unit values and export volumes, and interact the real exchange rate, as well as the real exchange rate interacted with quality, with a dummy for high (High) and a dummy for low (Low) income destination countries.50

The results for unit values are reported in Panel A of Table 10. According to column (1), higher quality is again associated with higher prices, and the coefficients on the real exchange rate, both for low and high income countries, are of similar magnitude. However, only the coefficient for high income countries is significantly different from zero so while pass-through is estimated as being complete for low income destinations, it decreases to approximately 90 percent for exports to high income countries. This indicates that price discrimination exists across destination countries and that the Law of One Price fails. This finding is also consistent with predictions from the literature that pricing-to-market should be stronger for higher income countries. For instance, Devereux, Engel, and Storgaard (2004) predict that a more stable monetary policy in high income destination countries reduces exchange rate pass-through by increasing the probability of invoicing in the currency of the destination country.

Table 10.

Heterogeneity across Destination Countries

article image
Notes: Year, province, firm-destination, grape, type, vintage year, and HS fixed effects are included in (1)-(4) while year-destination and product fixed effects are included in (5). Robust standard errors are adjusted for clustering at the product level. t-statistics in parentheses. a, b, and c indicate significance at 1, 5, and 10 percent levels, respectively. Unit values are in pesos per liter and export volumes are in liters.

Column (2) further interacts the exchange rate with quality and interestingly, the coefficient on the interaction term is positive and significant for high income destinations only. As a result, for low income countries, the response of unit values to exchange rate changes does not vary with quality. In contrast, for high income countries, the response of unit values to exchange rate changes increases with quality. A one standard deviation increase in quality increases the exchange rate elasticity from 0.112 to 0.120, i.e., a seven percent increase in pricing-to-market. These findings lend support to Prediction 3.

The finding that pass-through varies with quality for high income destination countries only is robust to the use of the number of destinations as a control for firm productivity (column 3), to restricting the sample to multi-product firms (column 4), and to the inclusion of destination-year and product fixed effects (column 5).

Panel B of Table 10 focuses on export volumes. Overall, the results find strong support for Prediction 4. Column (1) shows that a real depreciation raises export volumes to both low and high income destination countries. Consistent with the findings in Panel A, column (2) shows that the interaction between the real exchange rate and quality is significantly different from zero for high income countries only, and its negative sign further indicates that the response of export volumes to a change in the real exchange rate is smaller for higher quality wines. These findings remain robust in the other columns of the table, except when destination-year and product fixed effects are included in column (5).51

C. Aggregation Bias

As explained above, the degree of exchange rate pass-through that we find for the wine industry is consistent with the ones reported in other firm-level studies, but is larger than many estimates obtained using aggregate or industry-level data.52 Therefore, in this section we investigate whether pass-through estimates suffer from an aggregation bias.

To address this issue, we aggregate the data at different levels, re-estimate our benchmark specifications, and compare the magnitude of exchange rate pass-through estimated at each level of data aggregation. The first natural step would be to aggregate the data at the HS level, which classifies wines into five different categories. However, when doing so it turns out that the majority of firms actually export in a single HS code. We therefore instead aggregate the data at the firm-level (so firms become single-product) in which case the unit values capture the average price charged by each firm to each destination in each year, and quality is the average quality of all wines exported by each firm to each destination in each year, denoted by s¯ij,t,WS. In a second step, we further aggregate the data across firms in which case the unit values represent the average price of wine exported by all firms in the sample to each destination in each year, and quality is the average quality of all wine exports to each destination in each year and is denoted by s¯ij,t,WS.

Columns (1) and (2) of Table 11 replicate the benchmark specifications for unit values reported in columns (1) and (2) of Table 8. Whether or not we interact the real exchange rate with quality, average pass-through is estimated at 89 percent. Columns (3) and (4) of Table 11 report our two benchmark specifications estimated on the firm-level sample.53 As before, higher quality is associated with higher prices. The interaction between the exchange rate and quality in column (4) is however insignificantly different from zero, most likely because the averaging of quality across products at the firm level reduces the variation of quality in the sample substantially. Interestingly, in both columns average pass-through is now estimated at 69 percent, which is 20 percentage points lower than the 89 percent reported in columns (1) and (2). Columns (5) and (6) focus on the aggregate sample.54 Higher quality still leads to higher unit values although its significance is much reduced. In column (6) the interaction between the real exchange rate and quality is again insignificant,55 but it is interesting to note that average pass-through is low at 47 percent.

Table 11.

Aggregation Bias

article image
Notes: Year, province, firm-destination, grape, type, vintage year, and HS fixed effects are included in (1) and (2). Year, destination, and firm fixed effects are included in (3) and (4). Year and destination fixed effects are included in (5) and (6). Robust standard errors are adjusted for clustering at the product level in (1) and (2), firm level in (3) and (4), and destination level in (5) and (6). t-statistics in parentheses. a, b, and c indicate significance at 1, 5, and 10 percent levels, respectively. The real exchange rate qj,t elasticity (and therefore the magnitude of exchange rate pass-through) is evaluated at the mean value of quality in (2), (4) and (6).

To conclude, we show that part of the low exchange rate pass-through estimated using aggregate data appears to result from an aggregation bias: the more aggregated the data, the lower is estimated pass-through. In addition, the more aggregated the data, the more difficult it becomes to identify the impact of quality on heterogeneous pass-through.

VI. Robustness

In this section we discuss a number of alternative specifications we implement to ensure the robustness of our findings. Overall, the broad similarity of the resulting patterns is supportive of the paper’s main conclusions.56

A. The Measurement of Quality

We run a few sensitivity checks on the measurement of quality. Column (1) of Table 12 regresses equation (11) using the log of quality sWSk instead of its level. The results remain qualitatively unchanged, and as before a one standard deviation increase in quality from its mean level increases pricing-to-market by five percent.

Table 12.

Unit Values: Robustness on Quality

article image
Notes: The dependent variable is lnUVij,tk where unit values are in pesos per liter. Year, province, firm-destination, grape, type, vintage year, and HS fixed effects are included. Robust standard errors are adjusted for clustering at the product level. t-statistics in parentheses. a, b, and c indicate significance at 1, 5, and 10 percent levels, respectively. In (8) and (9), the instruments include the monthly average temperatures and total rainfall per province over the growing period (September to March) and the altitude of each province, and the same variables interacted with the exchange rate, respectively.

In order to minimize possible noise in the measurement of quality when defined on a (50,100) scale, we construct a new variable, denoted by s˜WSk, which takes on values between one and six where each value corresponds to one of the different bins defined by the Wine Spectator (see Table 5). A value of one indicates that the wine is “Not recommended” while a value of six that the wine is “Great” so a larger value captures a higher quality. The results of using s˜WSk as a regressor in (11) are reported in column (2) and remain qualitatively similar, although the magnitude of the estimated coefficient on quality becomes larger.

In columns (3) and (4), quality is measured using the Parker rankings and is denoted by sPk. The regression in column (3) includes all wines for which the Parker rankings are available, while column (4) restricts the sample to the wines for which both the Parker and the Wine Spectator rankings are simultaneously available. Qualitatively, our results largely hold up. The coefficient on sPk is however larger than the one on the Wine Spectator rankings.

Recall that due to missing observations on the Wine Spectator rankings, our sample covers 58 percent of the total FOB value exported by Argentina between 2002 and 2009. In order to increase the sample coverage, we calculate an average Wine Spectator ranking by wine name and type, and assign this average ranking to all wines with the same name and type. This increases our sample coverage to 85 percent of the total FOB value exported over the period. We apply this procedure to compute average quality on a (50,100) scale, denoted by sWSK, and on a (1,6) scale, denoted by s˜WSK where the K index indicates that quality varies by name and type only. The results of using either sWSK or s˜WSK are respectively reported in columns (5) and (6) of Table 12 and remain qualitatively unaffected.

Table 13 replicates the same specifications as in Table 12 but using export volumes as a dependent variable. Our results remain robust in all cases.

Table 13.

Export Volumes: Robustness on Quality

article image
Notes: The dependent variable is UVij,tk where export volumes are in liters. Year, province, firm-destination, grape, type, vintage year, and HS fixed effects are included. Robust standard errors are adjusted for clustering at the product level. t-statistics In parentheses. a, b, and c indicate significance at 1, 5, and 10 percent levels, respectively. In (8) and (9), the instruments include the monthly average temperatures and total rainfall per province over the growing period (September to March) and the altitude of each province, and the same variables interacted with the exchange rate, respectively.

B. The Endogeneity of Quality

One concern with our estimations is the potential endogeneity of quality in explaining unit values and export volumes. The Wine Spectator rankings are produced from blind tastings where the “price is not taken into account in scoring.” However, the “tasters are told […] the general type of wine (varietal and/or region) and the vintage” year.57 Similarly for Parker, “neither the price nor the reputation of the producer/grower affect the rating in any manner” although 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).”58 In other words, even if the two rankings are unaffected by the price, the tasters do have some basic information about the wines they taste which might in turn affect in a way or the other their scores, leading to an endogeneity bias which direction is, however, unclear. We therefore address the potential endogeneity of quality by using appropriate instruments.

The set of instruments we rely on to explain the variation in wine quality includes geographic and weather-related factors. Indeed, the literature devoted to explaining the quality of wine highlights that the amount of rainfall and the average temperatures during the growing season are strong determinants of quality (Ashenfelter, 2008; Ramirez, 2008). In the Southern hemisphere, the growing period spans the period from September (in the year before the vintage year) to March. In order to allow for the effects of temperature and rainfall to be nonlinear throughout the growing season, we consider as instruments the average temperature tp,m and the total amount of rainfall rp,m for each growing province p in each month m between September and March (Ramirez, 2008).59 Besides, one particularity of Argentina’s wine industry is the high altitude at which some of the growing regions are located, and there are strong reasons to believe that altitude contributes to variations in quality because it reduces the problems related to insects or grape diseases that affect quality at a low altitude. We therefore use the altitude Altp of each province p as an additional instrument for quality.60

The data on monthly average temperatures (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 Table 3 reports descriptive statistics on the average temperatures and total rainfall across growing regions. On average, temperatures are highest in January and lowest in September. January is also the wettest month and September the driest. Table 3 also shows that the provinces are on average 700 meters high, where altitude varies between 191 meters (province of La Pampa) and 1,238 meters (province of Salta).

As the instruments are only available over a reduced sample, we first replicate our benchmark OLS estimations reported in column (2) of Panels A and B of Table 8 for unit values and export volumes, respectively. The results, reported in column (7) of Tables 12 and 13 for prices and quantities, show that our main findings go through over the smaller sample.

Column (8) of Table 12 regresses by Instrumental Variables (IV) unit values on the real exchange rate, quality, and firm size. The coefficient on quality is positive and significant but becomes smaller compared to the OLS estimate in column (7). This positive endogeneity bias suggests that wine tasters tend to assign higher scores to more expensive wines. Column (8) of Table 13 focuses on export volumes. The instrumented effect of quality on export volumes is negative and significant, and is in turn larger in magnitude than the OLS estimate in column (7). For both regressions, the Kleibergen-Paap F statistic (equal to 44 for both the prices and quantities regressions, where the critical value is equal to 21, Stock and Yogo, 2005) largely rejects the null of weak correlation between the excluded instruments and the endogenous regressors.

The first-stage regressions for the two IV regressions (not reported due to space constraints but available upon request) show that climate variation affects wine quality. The results are somewhat erratic, but the positive coefficient on the February temperature is consistent with the finding in the literature that warmer temperatures during the harvest period (i.e., February/March in the Southern hemisphere) are typically associated with higher quality (the negative coefficient on the March temperature is therefore counterintuitive). Also, the positive coefficients on the October and December rainfall, and the negative coefficients on the January and February precipitations, are consistent with the expectation that precipitation during the earlier part of the growing season is good for quality, while a dry climate during the harvest period is more favorable for crops (Ramirez, 2008).

We then regress unit values and export volumes on quality which is further interacted with the exchange rate. The set of instruments for quality and for the interaction term now includes the monthly temperatures, monthly rainfall, and altitude variables as well as each of the variables interacted with the exchange rate.63 The results for unit values are reported in column (9) of Table 12 and show that exchange rate pass-through is larger for lower quality wines. Interestingly, the exchange rate elasticity increases from 0.105 to 0.131 following a one standard deviation increase in quality from its mean level, which corresponds to a 25 percent increase in pricing-to-market. This suggests that quality is quantitatively important in explaining heterogeneous pass-through. For export volumes in column (9) of Table 13, the coefficient on the interaction term is not statistically significant.

C. Asymmetries

We check if exporting firms adopt different pricing strategies depending on whether the Argentinean peso appreciates or depreciates in real terms. We estimate equations (11) and (12) for unit values and export volumes and interact the real exchange rate, as well as the real exchange rate interacted with quality, with a dummy for real appreciations (App) and a dummy for real depreciations of the peso (Dep).

Column (1) of Table 14 shows that the unit values response is larger when the peso appreciates than depreciates in real terms: the exchange rate elasticity evaluated at the mean value of quality is equal to 0.108 for real appreciations versus 0.092 for real depreciations (the two elasticities are statistically different from each other at the one percent level). This asymmetric pattern is consistent with firms trying to maintain export market shares by reducing more the domestic currency prices of their exports, which become less competitive when the peso appreciates. It is also consistent with the findings of Marston (1990) who shows that pricing-to-market by Japanese firms tends to be stronger in periods when the Japanese yen appreciates. Besides, column (1) of Table 14 also shows that a one standard deviation increase in quality from its mean level increases pricing-to-market by 8.3 percent following a real appreciation and by 4.3 percent only following a real depreciation. In turn, column (1) of Table 15 shows that for export volumes, a one standard deviation increase in quality reduces the exchange rate elasticity more during depreciations (decrease of 2.1 percent) than appreciations (decrease of 0.5 percent).

Table 14.

Robustness for Unit Values

article image
Notes: The dependent variable is ln UVij,tk, in (1)-(7) and (9) and UVij,mk in (7). Unit values are in pesos per liter in (1)-(8) and in USD per liter in (9). Year, province, firm-destination, grape, type, vintage year, and HS fixed effects are included. In (8), the year fixed effects are replaced by month fixed effects. Robust standard errors are adjusted for clustering at the product level. t-statistics in parentheses. a, b, and c indicate significance at 1, 5, and 10 percent levels, respectively.