Journal Issue
Share
Article

Quality and the Great Trade Collapse

Author(s):
Natalie Chen, and Luciana Juvenal
Published Date:
February 2016
Share
  • ShareShare
Show Summary Details

I. Introduction

One of the consequences of the global financial crisis has been a dramatic fall in world trade. The “Great Trade Collapse” occurred between the third quarter of 2008 and the second quarter of 2009, during which world trade fell by 30 percent in nominal terms, and 18 percent in real terms (World Trade Monitor, 2014). The crisis disproportionately affected trade in a narrow range of products, i.e., “postponable” goods, consumer goods, and investment goods (Baldwin, 2009). As these goods represent a large share of world trade (Engel and Wang, 2011), but a small fraction of world GDP, the world trade-to-GDP ratio collapsed dramatically.

The contribution of this paper is to explore empirically the effects of the global financial crisis on the composition of international trade. More precisely, the aim is to investigate whether the crisis has had heterogeneous effects on traded goods differentiated by quality. As higher quality goods are typically more income elastic than lower quality goods, we expect the adverse income shock induced by the crisis to have resulted in a “flight from quality” (Burstein, Eichenbaum, and Rebelo, 2005; Jaimovich, Rebelo, and Wong, 2015), whereby households in crisis-hit countries reduced not only the quantity, but also the quality of the goods they consume, leading to a sharper contraction in higher than in lower quality imports.1,2 We find strong evidence of a flight from quality in traded goods and, quantitatively, the effect is large: we estimate that up to nine percentage points difference in trade performance can be explained by the quality composition of traded goods.

From a theoretical point of view, the models of Fajgelbaum, Grossman, and Helpman (2011) and of Hummels and Lee (2012), which feature non-homothetic preferences, are particularly helpful in supporting our predictions.3Fajgelbaum and others (2011) show that as income falls, a smaller share of agents buys higher quality goods. Their result is therefore consistent with our expectation that trade in higher quality goods fell to a larger extent during the crisis. In Hummels and Lee (2012), under the assumption that the income elasticity of demand for higher quality goods is larger than for lower quality items, and that it is larger for poorer than for richer households, the model predicts that a negative income shock that hits most the left tail of the income distribution generates a sharper drop in the consumption of higher quality goods by poorer than by richer households.4 Aggregating over households results in the total consumption, and thus the imports, of higher quality goods to fall by more, consistent with a flight from quality driven by a negative income shock.

To establish whether the crisis has induced a flight from quality in traded goods, we focus our analysis on Argentinean wine exports, and investigate whether the exports of higher quality wines fell more dramatically during the crisis. Anecdotal evidence that the financial crisis had led to a flight from quality in wine consumption abounds. For instance, Veseth (2008) observes that during the crisis, “the $10 and up market segment” in the US was “still expanding, but the pace of growth had slowed considerably and there was evidence that buyers were trading down within it.” Thornton (2013) notes that between 2007 and 2009, US “consumers were paying a lower average price for the wine they purchased” as they moved away from the “higher-priced luxury wines they were consuming prior to the downturn, when they had higher incomes and there was less uncertainty about their employment status.” On Argentinean wines, Trad (2008) observes that “consumers were cutting back on wine spending” and were “willing to seek refuge in lower price ranges.”

To test our predictions, we use a rich and unique trade dataset from the Argentinean customs. The dataset reports, for each export flow between 2002 and 2009, the name of the exporting firm, the country of destination, the date of shipment, 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 rosé), and vintage year.5 To measure quality, we exploit two well-known experts wine ratings, the Wine Spectator and Robert Parker. Our approach to measuring quality builds on Chen and Juvenal (2015) where we use the same dataset to show that pricing-to-market increases with product quality, and is similar to Crozet, Head, and Mayer (2012) who match French firm-level exports of Champagne with experts quality assessments to study the relationship between quality and trade.

Our dataset is well-suited to study the heterogeneous response of traded goods differentiated by quality to the global financial crisis.6 First, as the level of disaggregation of the 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 traded goods (e.g., Behrens, Corcos, and Mion, 2013; Berthou and Emlinger, 2010; Levchenko, Lewis, and Tesar, 2011). As far as we know, our paper is the first to use such highly disaggregated product-level trade data to investigate the effects of the global financial crisis on the composition of international trade.7 Second, in contrast to papers that use trade unit values to proxy for quality, our focus on the wine industry allows us to rely on an observable measure of quality. When using the Wine Spectator scores that have the largest coverage of Argentinean wines, our sample includes 138 multi-product firms shipping 1,167 different wines with heterogeneous levels of quality. Aggregating our data at the 12-digit HS level would reduce our sample size by a factor of five, which would in turn significantly lower the within-firm variation in the quality of exported wines as the 138 exporters would only be selling at most three different “products.” Finally, export values are FOB and therefore measure the revenue received by exporters at the border, excluding nontradable factors such as transportation costs, tariffs, or distribution and retail costs in the importing country.

To understand the evolution of Argentinean wine exports, we perform a decomposition of nominal export growth and find that most of the adjustment during the crisis occurred at the intensive rather than at the extensive margin. As a result, in our empirical analysis we adopt a difference-in-difference specification which explains changes at the intensive margin, and compare the growth in export values, volumes, and unit values, before and during the crisis, across goods with different levels of quality. We identify the episode of the collapse in Argentinean wine exports by visually inspecting the data, and define the “crisis” period as running from the fourth quarter of 2008 to the third quarter of 2009 (henceforth, 2008Q4 and 2009Q3, respectively). The growth rates are computed for each firm-product-destination triplet that reports positive exports in two consecutive periods.

The central result of our paper is that before the crisis, higher quality goods enjoyed a stronger growth in nominal exports than lower quality products, but this trend reversed during the crisis and export growth fell more dramatically for higher quality goods. On average, a one unit increase on the quality scale raised export growth by two percentage points before the crisis, and reduced it by two percentage points during the downturn. The collapse in nominal exports for the higher quality wines was essentially driven by a drop in quantities, rather than a change in prices. This heterogeneity in the response of trade to the financial crisis remains robust to different measures of quality, samples, specifications, and to the potential endogeneity of quality.

Next, we investigate various channels to explain the sharper fall of higher quality exports during the crisis. First, we show that higher quality exports are more sensitive to changes in aggregate demand. Therefore, the negative income shock induced by the crisis was more detrimental to higher quality exports. Second, by splitting destination countries between wine producers and non-producers, and by distinguishing export markets based on their wine self-sufficiency, we show that the flight from quality was stronger in the countries where households had the option to substitute imports by domestically produced alternatives. Third, by relying on changes in the frequency of shipments at the transactions-level, we provide suggesting evidence that inventory adjustments matter to explain wine exports. However, as the effect turns out not to be heterogeneous across quality levels, we conclude that changes in inventories cannot explain the flight from quality during the crisis. Finally, we find that the flight from quality was stronger for the exports of smaller firms, which tend to be more sensitive to changes in aggregate demand and more credit constrained.

We also provide extensions to our benchmark results. First, we allow for nonlinearities in the effect of quality. The marginal effect of quality on trade, before and during the crisis, is significant for the higher quality wines only, and its magnitude rises with quality. Second, our results remain robust to running specifications in first differences, or in levels, that incorporate the extensive margin of adjustment. These regressions also show that the collapse in higher quality exports during the crisis was characterized not only by slower growth, but also by less exports. Finally, we extend the analysis post-crisis, and find that export growth recovered more strongly for higher quality wines, implying that the trade effects of the crisis were only temporary.

Finally, we assess the contribution of quality to the evolution of nominal and real wine exports. We use our regression estimates to evaluate how Argentinean wine exports would have performed during the crisis under two alternative scenarios. On the one hand, if the quality of all exported wines had increased during the crisis at the highest level observed in our dataset, nominal and real exports would have dropped by 38.94 and 40.63 percent, compared to an actual fall of 36.53 and 38.17 percent, respectively. On the other hand, if all ratings had instead fallen during the crisis at the lowest level of quality in the sample, nominal and real exports would have been reduced by 30.00 and 32.39 percent, respectively. The difference in nominal and real trade performance due to the quality composition of exports during the crisis is thus equal to nine and eight percentage points, respectively.

Our analysis concentrates on a specific sector in a single country. Evidence however suggests that quality heterogeneity matters more generally, and that differences in the quality of traded goods between countries and industries can be substantial (Feenstra and Romalis, 2014). Therefore, under the assumption that our findings generalize beyond the wine industry of Argentina, they carry macroeconomic implications. First, as they provide evidence that the composition of trade matters for the responsiveness of trade flows to economic downturns, they are helpful to infer how different countries are likely to perform in recessions. As higher income countries tend to be more specialized in the production of higher quality goods, our results imply that these countries’ exports might suffer more in recessions. Second, by showing that the crisis has impacted more the quantities than the prices of traded goods, our results suggest that the real effects of financial crises can be large. Finally, our findings have implications for understanding the distributional effects of crisis episodes as a flight from quality resulting from a negative income shock can be costly for consumers in terms of welfare. Under the assumption that consumers love variety, but also quality (Baldwin and Harrigan, 2011; Chen and Juvenal, 2015; Crozet and others, 2012; Johnson, 2012; Kugler and Verhoogen, 2012; Manova and Zhang, 2012b), a decline in the quality consumed reduces welfare.

Our paper belongs to several strands of the literature. The first one is the literature that addresses the role of quality during the recent trade collapse. Most closely related to our work are Bems and di Giovanni (2015), Berthou and Emlinger (2010), Esposito and Vicarelli (2011), and Levchenko and others (2011). We differ from these papers in several dimensions. First, we find evidence of a flight from quality in traded goods. Instead, Bems and di Giovanni (2015) document expenditure switching from imported to domestic goods. Berthou and Emlinger (2010) analyze prices data and conclude that a quarter of the decline in the EU import price during the crisis was driven by a lower demand for higher quality products. Esposito and Vicarelli (2011) estimate that the income elasticity of imports increases with quality, while Levchenko and others (2011) do not find any evidence that the imports of higher quality goods fell to a larger extent during the crisis. Second, we exploit the within-firm variation in quality of multi-product firms. In contrast, Berthou and Emlinger (2010), Esposito and Vicarelli (2011), and Levchenko and others (2011) rely on imports data at the HS or CN levels, while Bems and di Giovanni (2015) study scanner-data on the retail prices and quantities of products sold domestically. Third, our exports data are FOB and are thus free of insurance and transportation costs, while the import and retail data used by these papers are instead measured Cost, Insurance, and Freight (CIF), or in final consumer prices which include nontradables. Finally, our paper uses an observable measure of quality, while these papers rely on unit values as a proxy (Levchenko and others, 2011, also replicate the procedure of Khandelwal, 2010).8

Second, our analysis contributes to the literature that investigates changes in consumption patterns in the wake of aggregate shocks. In addition to Bems and di Giovanni (2015) mentioned above, evidence shows that during recessions, US households trade down in the quality of the goods and services they consume (Jaimovich and others, 2015), they reallocate their expenditures towards cheaper goods by switching to lower-price retailers (Coibion, Gorodnichenko, and Hong, 2015), and they purchase more on sale, use more coupons, buy larger sizes, and switch to generic products (Nevo and Wong, 2015).9Burstein and others (2005) show that the Argentinean devaluation of the early 2000s reduced the market shares of higher quality brands. Griffith, O’Connell, and Smith (2013) find that, during the global recession, UK households cut real expenditures on food, and substituted towards cheaper and less healthy food with lower nutritional quality.

Third, this paper relates to studies on the determinants of the Great Trade Collapse.10 Evidence suggests that trade contracted essentially because of a fall in aggregate demand (Baldwin, 2009), although Eaton, Kortum, Neiman, and Romalis (2015) argue that negative shocks to investment efficiency in durables are mostly to blame. Alessandria, Kaboski, and Midrigan (2010) emphasize the role of inventory adjustments, while Freund (2009) argues that trade is more sensitive to changes in GDP during downturns. Supply side explanations stress the role of cross-border vertical linkages (Bems, Johnson, and Yi, 2010, 2011; Levchenko, Lewis, and Tesar, 2010) and of financial constraints (Ahn, Amiti, and Weinstein, 2011; Bricongne, Fontagné, Gaulier, Taglioni, and Vicard, 2012; Chor and Manova, 2012). Evidence that trade protectionism increased during the crisis is limited (Eaton and others, 2015; Kee, Neagu, and Nicita, 2013). Regarding prices, Gopinath, Itskhoki, and Neiman (2012) find that in the US, the trade prices of non-differentiated manufactures sharply declined.11

Fourth, our work relates more broadly to papers that analyze trade patterns during macro-economic crises in general. Abiad, Mishra, and Topalova (2011) find that in countries hit by financial crises, imports decline sharply while exports are not affected. Amiti and Weinstein (2011) show that deteriorations in bank health during financial crises explain the drop in exports relative to output of Japanese firms. Bernard, Jensen, Redding, and Schott (2009) find that the intensive margin accounted for the majority of the changes in US trade around the 1997 Asian financial crisis.12

Finally, this paper contributes to the growing literature on quality and trade, which until recently mostly relied on trade unit values to measure quality.13Hummels and Klenow (2005) and Schott (2004) show that export unit values increase with exporter per capita income. Hallak (2006) and Hummels and Skiba (2004) find that richer countries have a stronger demand for high unit value exporting countries. Other papers investigate how quality relates to export performance using firm-level data (e.g., Kugler and Verhoogen, 2012; Manova and Zhang, 2012a,b; Verhoogen, 2008).14

The paper is organized as follows. Section 2 presents the firm-level customs data, the wine quality ratings, provides descriptive statistics, and investigates the relative contributions of the intensive and extensive margins to the collapse in Argentinean wine exports. Section 3 describes the empirical methodology, analyzes our main results, and explores various channels to explain the sharper fall of higher quality exports during the crisis. Section 4 discusses extensions. Section 5 addresses the economic significance of quality, while Section 6 provides robustness checks. Section 7 concludes.

II. Data and Descriptive Statistics

This section presents our data which come from two main sources: firm-level customs data and wine experts quality ratings. It also provides descriptive statistics, and investigates the relative contributions of the intensive and extensive margins to the collapse in Argentinean wine exports.

A. Firm-Level Customs Data

As in Chen and Juvenal (2015), the firm-level exports data are from the Argentinean customs and are obtained from a private vendor called Nosis. For each export flow we observe the name of the exporting firm, the destination country, the date of transaction, the 12-digit HS classification code, and the value (in US dollars) and volume (in liters) exported between 2002 and 2009.15 For each wine exported we have its name, type (red, white, or rosé), grape (Malbec, Chardonnay, etc.), and vintage year. Our sample covers a large range of destination countries that differ in terms of economic development, including OECD countries, EU countries, but also emerging markets (Brazil, India) and Asian countries (Hong Kong, Singapore). Figure 1, from Chen and Juvenal (2015), compares the total value of Argentinean wine exports from the customs dataset with the one provided by the Commodity Trade Statistics Database (Comtrade) of the United Nations (HS code 2204). The two series match extremely well.

Figure 1.Argentina’s Total Wine Exports

(million US dollars)

For details regarding the underlying database, and on how we clean up the raw data, see Chen and Juvenal (2015). We only keep the FOB flows and focus on wine producers (wholesalers and retailers are excluded, therefore each wine is exported by one firm only).16 Our dataset however differs from Chen and Juvenal (2015) in several respects. First, because we need to precisely date the episode of the trade collapse, we aggregate the data at a quarterly rather than at a yearly frequency.17 Second, we drop the shipments smaller than 4.5 liters (the latter corresponds to a carton of six 75cl bottles) to discard commercial samples exported for the purposes of marketing and promotion (including these flows tends to magnify the effect of quality on export growth). Third, one concern is that wine is an exhaustible resource: once a wine with a specific vintage year runs out, it can no longer be produced. To avoid the possibility that a wine becomes less exported during the crisis because its supply is running out, we define a product according to the name of the wine, its type, and grape, but ignore the vintage year (for each wine name, grape, type, destination, and time period, we sum exports across vintage years).18

As changes in nominal exports can be decomposed into changes in quantities and prices, we explore how the crisis has differentially impacted the growth in export values (in US dollars), volumes (in liters), and prices of goods with heterogeneous levels of quality. As export prices are not observed, as a proxy we compute the unit values of exports by dividing the value in US dollars by the volume in liters. As in Chen and Juvenal (2015), and in contrast to papers that define products according to trade classifications such as the CN or the HS (e.g., Behrens and others, 2013; Berthou and Emlinger, 2010; Levchenko and others, 2011), the granularity of our data ensures that compositional or quality changes do not affect movements in unit values (Feenstra and Romalis, 2014; Gopinath and others, 2012).

The literature typically assumes that the Great Trade Collapse started in 2008Q3 (which coincides with the bankruptcy of Lehman Brothers in September 2008) and ended around 2009Q2 (Behrens and others, 2013; Levchenko and others, 2011), with some variation across studies.19 For Argentina, we identify the episode of the trade collapse by visually inspecting the data. Figure 2 plots the year-on-year growth in quarterly Argentinean wine and total exports (in US dollars) between 2006Q4 and 2009Q4.20 Total exports fell by 10 percent in 2008Q4, witnessed a dramatic drop of 43 percent in 2009Q3, and started to recover in 2009Q4 (growth of -4.8 percent).

Figure 2.Year-on-year growth (%) in quarterly wine and total exports

(US dollars)

The collapse, and subsequent recovery, of total wine exports occurred in the same quarters. As shown in Figure 2, the year-on-year growth in quarterly wine exports dropped to 13.9 percent in 2008Q4, and continued to shrink during the first three quarters of 2009, declining by as a much as 9.8 percent in 2009Q3, and recovered in 2009Q4, reaching 2.4 percent. We therefore define the “crisis” period as running from 2008Q4 to 2009Q3, and for each firm-product-destination triplet we sum exports over this four-quarter period (henceforth, 2008Q4–2009Q3).21,22 We then calculate the growth in exports between two consecutive four-quarter periods, i.e., between the 2007Q4–2008Q3 and the 2008Q4–2009Q3 periods for the post-treatment (crisis) sample, and between the 2006Q4–2007Q3 and the 2007Q4–2008Q3 periods for the pre-treatment (pre-crisis) sample (Behrens and others, 2013; Iacovone and Zavacka, 2009). Computing changes between the same quarters avoid seasonality issues.23

Our sample spans the 2006Q4 to 2009Q3 time frame, and includes 5,478 wines of which 2,184 are exported between at least two (out of three) consecutive four-quarter periods. Once matched with the quality ratings, these numbers are reduced to 2,247 and 1,167 wines, which represent 59 and 50 percent of the total value of red, white, and rosé wine exported over the period. We sum export values and volumes across all wines and destinations in each four-quarter period, and calculate the growth rates of export values, volumes, and unit values. For the two samples including 2,247 and 1,167 wines, the collapse in trade during the crisis was substantial: nominal exports fell by 29.48 and 36.53 percent, real exports by 32.37 and 38.17 percent, while unit values rose by 4.11 and 2.57 percent, respectively. The growth in unit values could be, to some extent, attributed to internal factors taking place in Argentina, such as the rise in inflation or in financial costs (Gil-Fournier, 2008; Thornton, 2013).

B. Quality

To measure quality, we use the time-invariant quality scores of Chen and Juvenal (2015) which were collected from the Wine Spectator and Robert Parker magazines. The ratings are given on a (50,100) scale according to the name of the wine, its grape, type, and vintage year. A larger score indicates a higher quality.24Table 1 describes the Wine Spectator and Parker rating systems. For more details, see Chen and Juvenal (2015). As we define a wine product ignoring its vintage year, we compute, for each wine, the unweighted average of the Wine Spectator and Parker scores of all wines with the same name, grape, and type, and round it to its closest integer. As a robustness check, we also calculate a weighted average using the export shares of each wine as weights.

Table 1.Experts Ratings
Wine Spectator (50,100)Parker (50,100)
95-100Great96-100Extraordinary
90-94Outstanding90-95Outstanding
85-89Very good80-89Above average/very good
80-84Good70-79Average
75-79Mediocre60-69Below average
50-74Not recommended50-59Unacceptable
Notes: For both the Wine Spectator and Robert Parker rating systems, the table reports the quality bins corresponding to the scores assigned to each wine.
Notes: For both the Wine Spectator and Robert Parker rating systems, the table reports the quality bins corresponding to the scores assigned to each wine.

When we match the wines from the customs dataset with the average quality ratings from the Wine Spectator, we end up with 138 firms exporting 1,167 wines with 437 different names, three types, and 22 grapes. The lowest rated wine receives a score of 68, and the highest a score of 96. With Parker, we observe 595 wines exported by 107 firms (with 246 different names, three types, and 18 grapes). The average scores vary between 74 and 98 (only four out of the six bins listed in Table 1 are therefore included in the sample). We rely on the Wine Spectator for our main specifications because it has the largest coverage of Argentinean wines.

Figure 3 compares the Wine Spectator against the Parker ratings for the 487 wines, exported by 92 firms, reviewed by both tasters. For simplicity, quality is measured between one and six, where each value corresponds to one of the bins defined in Table 1. A value of one indicates that the wine is “Not recommended” or “Unacceptable,” and a value of six that the wine is “Great” or “Extraordinary,” but notice that only the three highest quality bins of Parker are plotted in the graph. The size of each circle is proportional to the number of wines at each point in the figure. The majority of wines are considered as “Above average/very good” (a value of four) by Parker, and as “Good” or “Very good” (a value of three or four) by the Wine Spectator. A relatively large number of wines are rated as “Outstanding” (a value of five) by both raters, and a few as “Great” or “Extraordinary” (a value of six). There is, however, also some disagreement between tasters as a few wines are rated as “Not recommended” (a value of one) by the Wine Spectator, but as “Above average/very good” (a value of four) by Parker. Still, the two ratings are positively correlated as Pearson’s correlation is equal to 0.49, and Kendall’s correlation index of concordance is 0.44. The figure is comparable to the one we report in Chen and Juvenal (2015) for the 2002 to 2009 period, and remains highly similar if plotted separately for the pre-crisis and crisis samples.25

Figure 3.Comparison between the Wine Spectator and Parker ratings

C. Descriptive Statistics

Table 2 shows that, with the exception of Brazil, Argentinean wine producers mostly exported to developed economies, the United States being the top destination market. Between the pre-crisis and crisis periods, the export shares to each destination market remained remarkably stable.

Table 2.Top Export Destinations
DestinationsExport share (%)

2007Q4–2008Q3
Export share(%)

2008Q4–2009Q3
United States15.9115.83
Netherlands7.347.42
United Kingdom7.286.44
Brazil4.734.16
Denmark4.483.12
Canada3.583.02
Finland2.082.25
Sweden2.052.80
Notes: The table reports total wine exports to each destination country as a share of total wine exports for the 2007Q4–2008Q3 and the 2008Q4–2009Q3 periods.
Notes: The table reports total wine exports to each destination country as a share of total wine exports for the 2007Q4–2008Q3 and the 2008Q4–2009Q3 periods.

Table 3 reports summary statistics for our main variables of interest, over the full sample and separately for the pre-crisis and crisis periods. Before the crisis, on average both nominal and real exports enjoyed a positive growth (of 6.55 and 3.43 percent), which subsequently collapsed with the downturn (by 7.15 and 12.12 percent, respectively). The mean growth in unit values instead remained steady over the whole period, and even accelerated during the crisis (from 3.12 to 4.97 percent). Besides, the mean quality ratings remained stable over time at a value of 84 for the Wine Spectator, and 88 for Parker.

Table 3.Summary Statistics
Full samplePre-crisisCrisis
Obs.MeanObs.MeanObs.Mean
Export values growth (%)8,0510.244,3446.553,707−7.15
Export volumes growth (%)8,051−3.734,3443.433,707−12.12
Unit values growth (%)8,0513.974,3443.123,7074.97
Wine Spectator8,051844,344843,70784
Parker5,138882,748882,39088
Number of firms8,0511384,3441263,707115
Number of wines8,0511,1674,3449963,707945
Number of destinations8,051844,344813,70776
Notes: For each variable listed in the first column, and separately for the full, pre-crisis, and crisis samples, the table reports the mean value and the number of observations available.
Notes: For each variable listed in the first column, and separately for the full, pre-crisis, and crisis samples, the table reports the mean value and the number of observations available.

Our full sample includes 138 exporters, 1,167 wines, and 84 destination countries (8,051 observations). The pre-crisis sample is composed of 126 firms exporting 996 wines to 81 countries (4,344 observations), while the crisis sample is slightly smaller, with 115 firms selling 945 wines to 76 export markets (3,707 observations). Among the 138 firms, export growth is observed in both periods for 103 exporters, and only before or during the crisis for 23 and 12 firms, respectively. The 35 firms exiting or entering the sample are small in terms of volumes exported (i.e., their total export quantities are below the 80th percentile of total real exports across firms, Bricongne and others, 2012), which is consistent with Castagnino, D’Amato, and Sangiácomo (2013) who document that small and medium-sized Argentinean firms are sporadic exporters. Export growth is observed in both periods for 774 wines, and only before or during the crisis for 222 and 171 wines, respectively. These wines are broadly similar in terms of quality as the mean Wine Spectator and Parker ratings are equal to 84 and 88 for the wines exported continuously, and to 84 and 89 for the ones entering or exiting from the sample. Finally, among the 84 export markets, 73 are served continuously, while eight drop out, and three enter, during the crisis.26 Overall, changes at the extensive margin for the number of firms, wines, and export markets are not negligible. However, as we show in Section 2.4, the collapse in wine exports during the crisis essentially occurred at the intensive margin. As a result, our main analysis explains changes at the intensive margin, while the extensive margin is addressed in Section 4.2.

Table 4 describes the data by quality bin of the Wine Spectator. For both sub-periods, “Good” and “Very good” wines represent the largest share of the sample (in terms of number of observations, firms, wines, destinations, and mean export shares). In contrast, “Great” and “Not recommended” wines have the smallest coverages. Higher quality wines are on average more expensive (Chen and Juvenal, 2015; Crozet and others, 2012). Due to the small number of observations for “Great” wines, we also merge “Great” and “Outstanding” wines into a single category. The same observations apply.

Table 4.Descriptive Statistics by Quality Bin of the Wine Spectator
ObservationsNumber of firms
Full samplePre-crisisCrisisFull samplePre-crisisCrisis
“Outstanding” and “Great”711403308433937
“Great”725212
“Outstanding”704401303433937
“Very good”3,0911,6721,419756960
“Good”3,7071,9731,734897776
“Mediocre”402217185373129
“Not recommended”1407961976
Number of winesNumber of destinations
Full samplePre-crisisCrisisFull samplePre-crisisCrisis
“Outstanding” and “Great”129105102585635
“Great”313424
“Outstanding”12610499585635
“Very good”353306289797573
“Good”533466431817872
“Mediocre”136105110343229
“Not recommended”161413292623
Mean unit valueMean export share (%)
Full samplePre-crisisCrisisFull samplePre-crisisCrisis
“Outstanding” and “Great”14.6014.1715.169.5810.438.32
“Great”22.3827.1720.470.020.010.03
“Outstanding”14.5214.1015.079.5610.428.29
“Very good”4.574.584.5539.8641.5037.47
“Good”4.244.154.3542.5639.8146.55
“Mediocre”3.493.063.995.906.345.27
“Not recommended”2.642.592.712.101.912.39
Notes: For each quality bin of the Wine Spectator, and separately for the full, pre-crisis, and crisis samples, the table reports the mean number of observations, exporters, wines, destinations, mean unit value (in US dollars per liter), and mean export share (%).
Notes: For each quality bin of the Wine Spectator, and separately for the full, pre-crisis, and crisis samples, the table reports the mean number of observations, exporters, wines, destinations, mean unit value (in US dollars per liter), and mean export share (%).

Table 5 describes the mean growth in export values, volumes, and unit values, before and during the crisis, by quality bin of the Wine Spectator. The table shows there is a significant amount of heterogeneity in the values taken by these variables across quality bins. Before the crisis, nominal export growth was on average the strongest in the higher quality segments, and reached 43.18 percent for “Great” wines, followed by 19.65 percent for “Outstanding,” and 11.02 percent for “Very good” wines. It was close to zero for “Good” wines, and equal to 3.01 and 6.39 percent for “Mediocre” and “Not recommended” wines. Interestingly, this pattern across quality bins reversed during the crisis. On average, the collapse in exports was the sharpest for “Great” wines at 31.88 percent, followed by 26.45 percent for “Outstanding,” 9.75 percent for “Very good,” 1.69 percent for “Good,” 4.49 percent for “Mediocre,” and 12.05 percent for “Not recommended” wines. Although indicative of a larger fall in trade for higher quality wines, these numbers remain silent on the behavior of exports within quality bins. They are however consistent with a flight from quality during the crisis. The pattern for export volumes is similar, while mean inflation dropped during the crisis for “Great” and “Outstanding” wines only. Similar observations apply if we merge “Great” and “Outstanding” wines into a single category.

Table 5.Mean Growth Rates by Quality Bin of the Wine Spectator
Export values (%)Export volumes (%)Unit values (%)
Pre-crisisCrisisPre-crisisCrisisPre-crisisCrisis
“Outstanding” and “Great”19.77−26.5313.97−19.235.79−7.30
“Great”43.18−31.8851.82−27.93−8.64−3.95
“Outstanding”19.65−26.4513.78−19.095.87−7.35
“Very good”11.02−9.757.69−15.393.335.64
“Good”0.45−1.69−2.87−7.413.325.72
“Mediocre”3.01−4.498.88−17.51−5.8613.02
“Not recommended”6.39−12.051.83−17.384.555.33
Notes: For each quality bin of the Wine Spectator, and separately for the pre-crisis and crisis samples, the table reports the mean growth in export values, volumes, and unit values.
Notes: For each quality bin of the Wine Spectator, and separately for the pre-crisis and crisis samples, the table reports the mean growth in export values, volumes, and unit values.

Finally, Table 6 describes mean export market shares. Using export values and volumes, we compute the share of exports of each wine in the total exports of all firms to each destination in each time period. We then report mean log changes by quality bin of the Wine Spectator. The value-based market shares of “Great” and “Outstanding” wines enjoyed the strongest mean growth before the crisis (of 49.41 and 17.58 percent), but decreased by more during the downturn (by 25.18 and 27.42 percent). A similar pattern applies to the volume-based market shares. These findings are again indicative of a flight from quality during the crisis. Similar conclusions obtain if “Great” and “Outstanding” wines are merged together.

Table 6.Mean Growth in Export Market Shares by Quality Bin of the Wine Spectator
Value-based shares (%)Volume-based shares (%)
Pre-crisisCrisisPre-crisisCrisis
“Outstanding” and “Great”17.73−27.3813.71−18.31
“Great”49.41−25.1859.42−15.85
“Outstanding”17.58−27.4213.49−18.35
“Very good”8.89−5.158.15−9.00
“Good”−0.670.67−0.14−3.52
“Mediocre”0.24−9.6310.02−20.38
“Not recommended”−1.56−0.390.65−2.78
Notes: For each quality bin of the Wine Spectator, and separately for the pre-crisis and crisis samples, the table reports the mean growth in the values- and volumes-based export market shares.
Notes: For each quality bin of the Wine Spectator, and separately for the pre-crisis and crisis samples, the table reports the mean growth in the values- and volumes-based export market shares.

D. Decomposition of Margins

To get a sense of the relative contributions of the extensive and intensive margins to the evolution of Argentinean wine exports during the crisis, we follow Behrens and others (2013) and Bernard and others (2009) and decompose nominal exports X in a given time period as X=f×d¯×p¯×x¯, where f denotes the number of exporting firms, d¯ the mean number of destination countries each firm exports to, p¯ the mean number of products each firm exports to each country, and x¯X/(f×d¯×p¯) are mean sales per firm-destination-product. Defining ΔX=X˜/X, where X˜ refers to exports in the following period, the change in exports from the 2007Q4-2008Q3 to the 2008Q4-2009Q3 period can be decomposed as

where Δf,Δd¯, and Δp¯ capture changes at the extensive margin while Δx¯ represents changes at the intensive margin, where the latter can be further decomposed into changes in mean quantities Δq¯ (in liters) and mean unit values Δuv¯ (in US dollars per liter), i.e., Δx¯Δq¯×Δuv¯. We perform this decomposition using the full sample of wine exports for which quality is observed (and which includes changes both at the intensive and at the extensive margin, i.e., with 2,247 wines).

As shown in Table 7, nominal wine exports fell by 29.48 percent during the crisis, and this collapse was driven by a fall in the number of exporters (by 5.33 percent), and in the mean number of destinations (by 6.84 percent) and of wines exported by each firm to each country (by 1.71 percent). Changes at the extensive margin therefore reduced exports by (0.9467 × 0.9316 × 0.9829 − 1) × 100 = −13.31 percent. The adjustment at the intensive margin however dominates as the mean value of exports per firm-destination-product fell by 18.65 percent. In other words, the relative contributions of the intensive and extensive margins to the collapse in wine exports amount to 63.27 and 36.73 percent, respectively.27 Changes at the intensive margin were essentially driven by a fall in the average quantities exported (by 21.99 percent), while mean unit values rose over the period (by 4.28 percent).28

Table 7.Decomposition of Margins
ExtensiveIntensive
Total exportsFirmsDestinationsProductsSalesQuantitiesPrices
2007Q4–2008Q3178,0711697.475.9823,5577,5453.12
2008Q4–2009Q3125,5851606.965.8819,1645,8863.26
Growth−29.48%−5.33%−6.84%−1.71%−18.65%−21.99%4.28%
Contribution36.73%63.27%
Notes: If we denote changes at the extensive and intensive margins as ΔEM and ΔIM, the contributions of the two margins to the total change in exports ΔX are given by ΔEMX ΔIMX. Total exports are in thousand US dollars and average sales per exporter-destination-product are in US dollars.
Notes: If we denote changes at the extensive and intensive margins as ΔEM and ΔIM, the contributions of the two margins to the total change in exports ΔX are given by ΔEMX ΔIMX. Total exports are in thousand US dollars and average sales per exporter-destination-product are in US dollars.

III. Quality and Trade During the Crisis

To determine whether the crisis has differentially impacted the growth in exports of goods with heterogeneous levels of quality, we estimate the following reduced-form regression (Behrens and others, 2013; Bernard and others, 2009; Iacovone and Zavacka, 2009)

where Δln Xijk,t is the log change of exports (in US dollars) of wine k exported by firm i to destination country j in period t. The log changes are calculated between the 2007Q4–2008Q3 and the 2008Q4–2009Q3 periods for the post-treatment (crisis) sample, and between the 2006Q4–2007Q3 and the 2007Q4–2008Q3 periods for the pre-treatment (pre-crisis) sample. As computing log changes requires us to observe positive trade flows for each firm-product-destination triplet between two consecutive four-quarter periods, equation (2) explains changes at the intensive margin which, as we have shown previously, dominated the evolution of wine exports during the crisis. The quality of wine k is denoted by qualityk, and the Wine Spectator ratings are used for our benchmark specifications. We define a dummy variable Dcrisis that is equal to one for the 2008Q4–2009Q3 crisis period, and interact it with quality. The coefficients to be estimated are β1 and β2, and εijk,t is an error term. The coefficient β1 captures the effect of quality on export growth in a “normal” period (i.e., before the crisis), while β2 measures the differential effect of quality on export growth during the crisis.

We control for an extensive set of fixed effects and perform within estimations. We include firm-destination-time effects, Dij,t, that sweep out all aggregate, firm, and destination-specific supply and demand shocks that are common across the goods exported by each firm to each destination country at each point in time. These include factors that vary by firm-destination-time (e.g., 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 country), time-varying characteristics of the exporters such as changes in productivity, firm size, global value chains, inventories, or credit constraints, firm-destination effects such as the tastes of an importer for the goods of a firm, and time-varying destination-specific factors such as GDP growth, multilateral resistance, protectionist measures, or bilateral exchange rates.29 They also absorb the direct effect of the crisis on Argentinean wine exports (i.e., the main effect of Dcrisis). As product fixed effects are collinear with quality, we control for product characteristics by including grape Dgrape, type Dtype, HS DHS, and province of origin of the grapes Dp fixed effects.30 Controls for the wine names are not included as they are collinear with the firm fixed effects (as each wine is exported by one firm only). Standard errors are clustered by destination-time to control for time-varying idiosyncratic shocks correlated at the importer-level.

As our main interest lies in the interaction between quality and the crisis dummy variable, we proceed to estimate a more conservative specification

which controls for product fixed effects, Dk. As a result, quality drops out from the regression. The coefficient of interest, γ, captures the differential impact of quality on export growth during the crisis.

In addition to export values, we also estimate equations (2) and (3) using the log change of export volumes and unit values as dependent variables. Given that unit values are equal to the ratio between values and volumes, the elasticities for export values are, by construction, equal to the sum of the elasticities for export volumes and unit values.

A. Baseline Results

Panel A of Table 8 reports the results for nominal exports. In column (1), we estimate equation (2) but only include quality as a regressor, which is insignificant. Once we let the relationship between export growth and quality vary over time, column (2) shows that higher quality wines enjoyed a stronger growth in exports before the crisis, but were subsequently more negatively affected. On average, a one unit increase in quality raised export growth by two percentage points before the crisis, and reduced it by two percentage points during the downturn (0.018 – 0.036). The results are similar in column (3), where the grape, type, province, and HS fixed effects are interacted with time dummies to control for unobserved, time-varying product characteristics (for instance, white and rosé wines mature for a shorter period of time compared to red wines, and need to be sold within a few months after fermentation, Thornton, 2013), or for shifts in consumer preferences for different types of wines.

Table 8.Baseline Results
(1)(2)(3)(4)(5)(6)(7)
Panel A: Export values
quality0.0020.018b0.017b
(0.007)(0.008)(0.008)
quality×Dcrisis−0.036a−0.034a−0.030b−0.027b−0.044c−0.030b
(0.013)(0.012)(0.014)(0.012)(0.026)(0.014)
R-squared0.5010.5020.5040.5920.5770.6290.565
Panel B: Export volumes
quality0.0060.020b0.019b
(0.007)(0.008)(0.009)
quality×Dcrisis−0.031b−0.029b−0.026b−0.023b−0.043c−0.026b
(0.013)(0.012)(0.013)(0.012)(0.026)(0.013)
R-squared0.4950.4960.4980.5880.5800.6160.563
Panel C: Unit values
quality−0.004a−0.002−0.002
(0.002)(0.002)(0.002)
quality×Dcrisis−0.005c−0.005−0.004−0.004c−0.001−0.004
(0.003)(0.003)(0.002)(0.002)(0.006)(0.002)
R-squared0.5070.5080.5110.6000.5720.6520.564
SampleFullFullFullFullFullFullInt k
QualityWSWSWSWSWSParkerWS
Wine characteristics*YesYesYes NoNoNoNo
Product FENoNoNoYesYesYesYes
WeightedNoNoNoNoYesNoNo
Observations8,0518,0518,0518,0518,0515,1387,505
Notes: Firm-destination-time, grape, type, province, and HS fixed effects are included in (1) and (2). The grape, type, province, and HS fixed effects are interacted with time dummies in (3). Product fixed effects are included in (4) to (7). Weighted regression in (5) where the weights are the lagged values of export volumes in liters. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b and c indicate significance at the one, five, and ten percent levels. * “Wine characteristics” refers to grape, type, province, and HS fixed effects. † “Yes” indicates that the wine characteristics fixed effects are further interacted with time dummies.
Notes: Firm-destination-time, grape, type, province, and HS fixed effects are included in (1) and (2). The grape, type, province, and HS fixed effects are interacted with time dummies in (3). Product fixed effects are included in (4) to (7). Weighted regression in (5) where the weights are the lagged values of export volumes in liters. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b and c indicate significance at the one, five, and ten percent levels. * “Wine characteristics” refers to grape, type, province, and HS fixed effects. † “Yes” indicates that the wine characteristics fixed effects are further interacted with time dummies.

The results of estimating equation (3) are reported in column (4). Quality interacted with the crisis dummy remains negative and significant, albeit slightly smaller in magnitude. To make sure that our results are not driven by observations that are economically small, column (5) weights the observations by the volume of exports in the previous period (Behrens and others, 2013). In column (6), quality is measured using the Parker ratings. Finally, column (7) only includes the wines exported both before and during the crisis to any destination. Qualitatively, our results hold in all cases.31

Panels B and C report the same specifications for export volumes and unit values, respectively. For volumes, the pattern and magnitude of the results are very similar to the ones for export values. In contrast, for unit values the interactions between quality and the crisis dummy are negative, but mostly insignificant. The collapse in exports of higher quality wines during the crisis was therefore driven by a drop in quantities, rather than a change in prices (Behrens and others, 2013).

B. The Channels

This section explores various channels to explain the sharper fall of higher quality exports during the financial crisis. We consider the role of changes in aggregate demand, expenditure switching from imported to domestically produced varieties, inventory adjustments, and firm size.32.

1. Aggregate Demand

Our presumption is that the flight from quality that we identify in the data was primarily driven by a fall in aggregate demand, combined with higher quality goods being more income elastic than lower quality items. To investigate this channel, we control for real GDP per capita growth in each destination country to capture changes in aggregate demand. We also use the growth in household real final consumption expenditures per capita, and in the unemployment rate, as alternative business cycle indicators (due to incomplete country coverage, the sample sizes are slightly reduced).33

We estimate equation (3), but only include quality interacted with the growth of each aggregate demand proxies as a regressor (the main effects of aggregate demand and of quality being absorbed by the fixed effects). The results for export values are reported in Panel A of Table 9. In column (1), the interaction between quality and income growth is positive and significant. In other words, the drop in mean income growth that we observe in the sample, from 1.6 to -3.9 percent during the crisis, was more detrimental to higher quality exports. The results remain consistent if we measure changes in aggregate demand using the growth in household expenditures per capita (column 2), which on average fell from 1.6 to -2.0 percent during the crisis, or in the unemployment rate (column 3), which on average rose from -0.1 to 1.9 percent (in column 3, the negative interaction indicates that the rise in unemployment during the crisis reduced higher quality exports to a larger extent). The results for volumes in Panel B are similar. For unit values in Panel C, the interactions are insignificant.34

Table 9.Aggregate Demand
(1)(2)(3)(4)(5)(6)
Panel A: Export values
Δ ln AD×quality0.411a0.372b−0.073a0.340b0.302c−0.057c
(0.139)(0.159)(0.028)(0.155)(0.179)(0.031)
R-squared0.5930.5960.5920.6630.6700.664
Panel B: Export volumes
Δ ln AD×quality0.397a0.377b−0.068a0.364b0.329c−0.059b
(0.133)(0.163)(0.025)(0.145)(0.182)(0.028)
R-squared0.5890.5930.5890.6600.6680.662
Panel C: Unit values
Δ ln AD×quality0.014−0.005−0.005−0.024−0.0270.002
(0.028)(0.031)(0.006)(0.039)(0.034)(0.009)
R-squared0.6010.6060.5990.6680.6760.667
Aggregate demand ADGDP pcExp. pcUnempl.GDP pcExp. pcUnempl.
Product-time FENoNoNoYesYesYes
Observations7,9907,4927,7417,9907,4927,741
Notes: Firm-destination-time and product fixed effects are included in (1) to (3). Product-time fixed effects are further included in (4) to (6). Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.
Notes: Firm-destination-time and product fixed effects are included in (1) to (3). Product-time fixed effects are further included in (4) to (6). Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.

Columns (4) to (6) include product-time fixed effects to control for unobserved, time-varying characteristics of the different wines (such fixed effects could not be included in the regressions reported in Table 8 as they are collinear with the main variable of interest, i.e., quality interacted with the crisis dummy variable). For nominal and real exports, the interactions between quality and changes in aggregate demand remain significant and with expected signs.

Another way to explore the role of aggregate demand is to focus on specific destination countries. Indeed, the global financial crisis has hit some countries more than others, with some countries falling deeply into recession and others avoiding it. The resulting drop in aggregate demand, and in turn in the demand for Argentinean wines, can thus be expected to be stronger in some countries than in others. Examples include the US (the epicenter of the global financial crisis) and the UK, where real GDP per capita growth fell from -0.8 and -0.6 percent to -5.2 and -5.9 percent during the crisis, respectively.35 Conversely, as China has remained immune to the global recession (with a real income growth of 8.5 and 8.8 percent before and during the crisis), higher quality wine exports to China can be expected to have grown.36,37 Finally, Hong Kong entered into a recession in 2008, with income per capita growth shrinking from 1.2 to -4.2 percent. At the same time, however, its ad-valorem import duties on wine were cut from 40 to zero percent in February 2008, and all import licences and permits on wine were removed in June 2008, potentially stimulating higher quality wine imports.38 For each country, we therefore investigate how the exports and prices of higher quality goods performed during the crisis relative to lower quality items. We estimate equation (3), and let the interaction between quality and the crisis indicator vary between each country and the other countries in the sample.39

Table 10 shows that higher quality wine exports to the US and the UK fell to a large extent during the crisis. In nominal terms (Panel A), a one unit increase in quality reduced export growth from its pre-crisis level by 11.2 percentage points for the US (column 1), and by 7.3 percentage points for the UK (column 2), compared to around 2.5 percentage points for the other countries. Both the quantities (Panel B) and the prices (Panel C) of higher quality wine exports decreased by more for the US, while for the UK only the quantities declined more sharply. For China, the interactions for nominal and real exports are insignificant (column 3). Despite the sustained growth that China has enjoyed during the crisis, its imports were not differentially affected across quality levels, although higher quality wines experienced lower inflation. For Hong Kong, a one unit increase in quality raised the growth rates of nominal and real exports from their pre-crisis levels by 7.0 and 7.8 percentage points, respectively, and reduced the growth rate of unit values by 0.8 percentage points (column 4).40 The impact of the liberalization measures introduced during the crisis has thus offset the negative trade effect attributable to the fall in aggregate demand. Column (5) controls for the different countries simultaneously, and the results remain similar.

Table 10.Destination Countries
(1)(2)(3)(4)(5)
Panel A: Export values
quality×Dcrisis × others−0.024c−0.026c−0.030b−0.032b−0.021c
(0.014)(0.014)(0.014)(0.014)(0.012)
quality×Dcrisis × US−0.112a−0.114a
(0.023)(0.024)
quality×Dcrisis × UK−0.073a−0.076a
(0.023)(0.022)
quality×Dcrisis × China0.0090.008
(0.031)(0.031)
quality×Dcrisis × Hong Kong0.070a0.072a
(0.018)(0.018)
R-squared0.5930.5920.5920.5920.593
Panel B: Export volumes
quality×Dcrisis × others−0.020−0.023c−0.027b−0.028b−0.018
(0.013)(0.013)(0.013)(0.013)(0.012)
quality×Dcrisis × US−0.100a−0.102a
(0.022)(0.023)
quality×Dcrisis × UK−0.069a−0.072a
(0.021)(0.019)
quality×Dcrisis × China0.0360.035
(0.034)(0.034)
quality×Dcrisis × Hong Kong0.078a0.080a
(0.017)(0.017)
R-squared0.5890.5890.5890.5890.589
Panel C: Unit values
quality×Dcrisis × others−0.003−0.004−0.004−0.004−0.003
(0.002)(0.002)(0.002)(0.002)(0.003)
quality×Dcrisis × US−0.012a−0.012a
(0.004)(0.004)
quality×Dcrisis × UK−0.004−0.004
(0.004)(0.004)
quality×Dcrisis × China−0.027b−0.027b
(0.012)(0.012)
quality×Dcrisis × Hong Kong−0.008a−0.008a
(0.003)(0.003)
R-squared0.6000.6000.6000.6000.600
Observations8,0518,0518,0518,0518,051
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.

2. Expenditure Switching

In response to an adverse income shock, households reduce both the quantity and the quality of the goods they consume, leading to a sharper contraction in higher than in lower quality imports. Our expectation is that this flight from quality in traded goods should be stronger in the countries where households have the option to substitute imports by domestic alternatives. Such a substitution could happen because imports are more expensive than domestic goods (Alchian and Allen, 1964), or because households tend to buy relatively more from home country suppliers during downturns as they become more patriotic in product choices or have stronger trust in domestic goods.41 Using data on the retail prices and quantities of products sold in Latvia, Bems and di Giovanni (2015) find evidence of a substitution from imported to domestically produced goods in response to the 2008–2009 balance of payments crisis.

Missing any data on the consumption and the prices of domestically produced wines with different and comparable levels of quality in each destination country, we are unable to directly assess expenditure switching from imported to domestically produced wines. Using data on total wine production (Food and Agriculture Organization of the United Nations), an alternative way of proceeding is to split the sample between wine producing and non-producing countries, and to check whether higher quality Argentinean wine exports fell by more when shipped to wine producing destinations. Indeed, expenditure switching from imports to domestic goods is only possible in wine producing countries.42

Table 11 summarizes the mean growth in the values, volumes, and unit values of wines exported to wine producing and non-producing countries. Before the crisis, nominal and real exports enjoyed a positive mean growth to both producing (of 6.06 and 2.61 percent) and non-producing nations (of 7.49 and 5.04 percent), but this collapsed with the onset of the crisis. Interestingly, nominal and real exports fell by 10.10 and 16.26 percent on average to wine producing countries, compared to only 0.76 and 3.15 percent to non-wine producers.

Table 11.Mean Growth Rates for Wine Producing and Non-Producing Countries
Export values (%)Export volumes (%)Unit values (%)
Pre-crisisCrisisPre-crisisCrisisPre-crisisCrisis
Wine non-producing countries7.49−0.765.04−3.152.452.38
Wine producing countries6.06−10.102.61−16.263.466.16
Largest five−2.52−26.45−5.69−30.073.163.62
Largest ten−1.60−8.13−5.15−15.513.557.38
Largest fifteen1.19−8.79−2.36−15.553.556.76
Notes: For each group of countries listed in the first column, and separately for the pre-crisis and crisis samples, the table reports the mean growth in export values, volumes, and unit values.
Notes: For each group of countries listed in the first column, and separately for the pre-crisis and crisis samples, the table reports the mean growth in export values, volumes, and unit values.

As expenditure switching is likely to be weaker in the nations that produce little wine (e.g., Canada), Table 11 also summarizes our data for the five, ten, and 15 major producing countries. On average, nominal and real export growth was negative before the crisis (with the exception of nominal growth to the 15 largest producers), but fell considerably more during the downturn, especially to the five largest producers. Nominal export growth to these countries fell on average by 26.45 percent, versus 8.13 and 8.79 percent to the ten and 15 largest producing nations, respectively. Besides, for all groups of producing countries, mean export inflation was higher during the crisis than before.

We then estimate equation (3), and let quality interacted with the crisis dummy vary between three groups of countries: the largest producers, the other producers, and the non-producers, where the major producers are in turn defined as the five, ten, and 15 largest producing countries.43 In column (1) of Panel A in Table 12, the interaction is large and negative at a value of -0.087 for the five major producers, equals -0.027 for the other producers, and is insignificant for the non-producers. A one unit increase in quality therefore reduced nominal export growth from its pre-crisis level by 8.7 percentage points to the five largest producing nations, versus 2.7 percentage points to the others. A similar pattern is observed in columns (2) and (3) for the ten and 15 largest producers. Notice that the magnitude of the interaction term for the largest producers decreases gradually as we move from column (1) to (3). Similar observations apply for export volumes in Panel B and unit values in Panel C. Despite becoming cheaper in relative terms, higher quality wines therefore suffered more dramatically when exported to the largest producing countries.

Table 12.Wine Producing and Non-Producing Countries
(1)(2)(3)(4)(5)(6)
Panel A: Export values
quality×Dcrisis × largest−0.087a−0.069a−0.052b−0.060a−0.023−0.034c
(0.021)(0.023)(0.020)(0.021)(0.035)(0.019)
quality×Dcrisis × non – largest−0.027c−0.027−0.027−0.015−0.021−0.013
(0.016)(0.017)(0.020)(0.016)(0.017)(0.020)
quality×Dcrisis × non – producers−0.017−0.017−0.0180.0030.0030.004
(0.016)(0.016)(0.016)(0.019)(0.019)(0.020)
R-squared0.5930.5930.5920.5930.5930.593
Panel B: Export volumes
quality×Dcrisis × largest−0.073a−0.056b−0.041b−0.045b−0.007−0.021
(0.021)(0.023)(0.020)(0.021)(0.036)(0.019)
quality×Dcrisis × non – largest−0.023−0.025−0.026−0.010−0.017−0.011
(0.015)(0.016)(0.019)(0.015)(0.015)(0.019)
quality×Dcrisis × non – producers−0.016−0.016−0.0160.0060.0060.006
(0.017)(0.017)(0.017)(0.019)(0.018)(0.019)
R-squared0.5890.5890.5890.5890.5900.589
Panel C: Unit values
quality×Dcrisis × largest−0.013a−0.013a−0.011a−0.015a−0.017a−0.013a
(0.004)(0.004)(0.003)(0.005)(0.005)(0.004)
quality×Dcrisis × non – largest−0.003−0.003−0.001−0.006−0.004−0.002
(0.003)(0.003)(0.003)(0.004)(0.004)(0.004)
quality×Dcrisis × non – producers−0.001−0.001−0.001−0.003−0.003−0.003
(0.004)(0.004)(0.004)(0.005)(0.005)(0.005)
R-squared0.6000.6000.6000.6010.6020.602
Wine producing countries largestFiveTenFifteenFiveTenFifteen
GDP per capitaNoNoNoYesYesYes
Observations8,0518,0518,0517,9907,9907,990
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.

With the exception of China, the five major producing countries (France, Italy, Spain, and the United States) were all severely hit by the crisis. To ensure that our results are not driven by changes in aggregate demand, columns (4) to (6) include real GDP per capita growth interacted with quality, separately for the three groups of countries (not reported). The coefficients on quality interacted with the crisis indicator become smaller in magnitude and less significant, but our results still hold.

Another indirect way to investigate expenditure switching is to distinguish countries based on their wine self-sufficiency. Self-sufficiency is given by the volume of domestic wine production over consumption, and captures whether a country is dependent, or not, on imports to satisfy domestic demand. It is equal to zero for non-producing countries, and is larger or smaller than one for self-sufficient and non-self-sufficient nations, respectively.44 In response to a negative income shock, our expectation is that the flight from quality should be more acute in self-sufficient countries, where households can substitute imports by domestic alternatives which supply is large enough to satisfy local demand. We compute a dummy variable for the non-producing, self-sufficient, and non-self-sufficient countries, and let quality interacted with the crisis indicator in equation (3) vary between the three country groups.

Table 13 reports the results. For export values in column (1), quality interacted with the crisis dummy is again insignificant for the non-producing nations. It is instead negative at a value of -0.026 for the non-self-sufficient countries, and increases in magnitude at a value of -0.073 for the self-sufficient destinations. For export volumes and unit values (columns 2 and 3), the interaction is significant for self-sufficient countries only. Column (4) to (6) control for growth in income per capita interacted with quality, separately for the three groups of countries, and qualitatively the results remain similar.45

Table 13.Self Sufficiency
Dependent variableExportExportUnitExportExportUnit
valuesvolumesvaluesvaluesvolumesvalues
quality×Dcrisis × sufficient–0.073a–0.064a–0.009b–0.072a–0.060b–0.011c
(0.023)(0.022)(0.004)(0.024)(0.024)(0.006)
quality×Dcrisis × non – sufficient–0.026c–0.022–0.004–0.010–0.004–0.006
(0.016)(0.016)(0.003)(0.017)(0.015)(0.004)
quality×Dcrisis × non – producers–0.018–0.016–0.0010.0030.006–0.003
(0.016)(0.017)(0.004)(0.019)(0.019)(0.005)
R-squared0.5930.5890.6000.5930.5890.601
GDP per capitaNoNoNoYesYesYes
Observations8,0518,0518,0517,9907,9907,990
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.

3. Inventory Adjustments

Alessandria and others (2010) claim that inventory adjustments contributed to the trade collapse during the global financial crisis. In response to a negative demand shock, importers may indeed deplete their inventories instead of importing more from abroad, amplifying the magnitude of the trade decline. One way to address the role of inventories is to consider, for each wine exported to each destination country in each time period, the change in the frequency of shipments at the transactions-level as a lower frequency could be an indication that the destination country is depleting its inventories.

To check if inventory adjustments explain the growth of Argentinean wine exports, and of higher quality wines in particular, equation (3) includes the log change in the frequency of shipments, on its own and interacted with quality, as regressors (product-time fixed effects are controlled for). In columns (1) to (3) of Table 14, the change in frequency is positive and significant, as expected, but its interaction with quality is insignificant for export values and volumes, and negative for unit values.

Table 14.Inventory Adjustments
(1)(2)(3)(4)(5)(6)
Dependent variableExportExportUnitExportExportUnit
valuesvolumesvaluesvaluesvolumesvalues
Δ ln frequency1.589b1.245c0.343b
(0.781)(0.750)(0.165)
Δ ln frequency × quality–0.006–0.002–0.004b
(0.009)(0.009)(0.002)
Δ ln frequency × red1.742b1.319c0.423b
(0.781)(0.21)(0.19)
Δ ln frequency × white/rosé0.9110. 80.124
(1.564)(1.551)(0.249)
Δ ln frequency × quality × red–0.008–0.003–0.005b
(0.009)(0.008)(0.002)
Δ ln frequency × quality × white/rosé0.0020.003–0.001
(0.018)(0.018)(0.003)
R-squared0 7660.7730.6670.7660.7740.667
Observations8,0518,0518,0518,0518,0518,051
Notes: Firm-destination-time and product-time fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.
Notes: Firm-destination-time and product-time fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.

As white and rosé wines need to be sold within a few months after fermentation, inventory adjustments could matter more for red wines as the latter can be stored. In columns (4) to (6) we therefore let the frequency of shipments, and its interaction with quality, vary between red and white/rosé wines. Frequency is positive and significant for red wines only, but its interaction with quality remains insignificant (and negative for unit values). We therefore conclude that inventory adjustments matter to explain trade flows, but as their effect is not heterogeneous across goods differentiated by quality, they cannot explain the flight from quality that we observe during the crisis.

4. Firm Size

This section investigates whether the trade effects of the crisis vary between exporters of different sizes. We hypothesize that the flight from quality should be stronger for the exports of smaller firms. First, if importers anticipate to sell less higher quality wines due to the recession, they may cut down on their purchases from the small wineries and instead continue to import from the largest producers who are more important business partners to them. Second, on the supply side, as small firms are typically more restricted in their access to external finance than large firms, the tightening of credit conditions during the crisis could reduce the exports of small firms by more (Bricongne and others, 2012; Chor and Manova, 2012).46,47 Lack of external financing could also be more detrimental to higher quality wine exports as the latter face higher (per unit) transportation and distribution costs (Chen and Juvenal, 2015) which need to be incurred before export revenues are realized, and which payments are therefore highly reliant on external credit availability.

To measure firm size, we rank firms according to their total volume of exports (in liters), and classify them into two categories. As in Bricongne and others (2012), based on the 80th percentile of total exports, the small group includes the smallest 80 percent of exporters, while the large group includes the others. We estimate equation (3), and let quality interacted with the crisis dummy vary by firm size. Consistent with expectations, in both nominal and real terms the exports of higher quality wines collapsed to a larger extent during the crisis for the small firms, while no significant effect can be detected for the large firms (columns 1 and 2 of Table 15). The results for unit values are insignificant (column 3). Columns (4) to (6) also include income growth interacted with quality separately for each firm size. For nominal exports (column 4), the coefficient for the small firms is reduced in magnitude but remains significant, suggesting that demand side factors are not the only reason for the collapse of these firms’ higher quality exports, and that supply side factors, such as tighter credit conditions, may have also played a role (although the results for unit values are insignificant).

Table 15.Firm Size
(1)(2)(3)(4)(5)(6)
Dependent variableExportExportUnitExportExportUnit
valuesvolumesvaluesvaluesvolumesvalues
quality×Dcrisis × small−0.048a−0.042b−0.006−0.042c−0.033−0.009
(0.019)(0.017)(0.004)(0.022)(0.020)(0.007)
quality×Dcrisis × large−0.018−0.015−0.0020.0040.007−0.003
(0.016)(0.017)(0.004)(0.017)(0.016)(0.005)
R-squared0.5920.5890.6000.5930.5890.601
GDP per capitaNoNoNoYesYesYes
Observations8,0518,0518,0517,9907,9907,990
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.

IV. Extensions

This section discusses extensions to our benchmark specifications. We investigate nonlinearities in the effect of quality, extensive margin adjustments, and the post-crisis recovery.

A. Nonlinearities

To gain further insights on the relationship between trade and quality, we investigate whether the marginal effect of quality is heterogeneous across goods with different qualities. We estimate equation (2) and include quality and its squared value, on their own and interacted with the crisis dummy. Columns (1) to (3) of Table 16 show that nonlinearities are present for export values, volumes, and (less so) unit values. The negative squared quality terms during the crisis indicate that the semi-elasticities of export values and volumes to quality increase in magnitude at higher levels of quality.

Table 16.Nonlinearities
(1)(2)(3)(4)(5)(6)
Dependent variableExportExportUnitExportExportUnit
valuesvolumesvaluesvaluesvolumesvalues
quality−0.227−0.2590.032
(0.185)(0.189)(0.037)
quality×Dcrisis0.518b0.473b0.0440.3510.3420.009
(0.219)(0.217)(0.051)(0.236)(0.243)(0.060)
quality squared0.0010.0020.000
(0.001)(0.001)(0.000)
quality squared×Dcrisis−0.003b−0.003b0.000−0.002−0.0020.000
(0.001)(0.001)(0.000)(0.001)(0.001)(0.000)
Semi-elasticities (pre-crisis)
“Great”0.061b0.069b−0.008
(0.031)(0.031)(0.008)
“Outstanding”0.045b0.051a−0.006
(0.019)(0.019)(0.005)
“Very good”0.030a0.034a−0.004
(0.010)(0.010)(0.003)
“Good”0.015c0.017c−0.002
(0.009)(0.009)(0.002)
“Mediocre”0.0010.0010.001
(0.017)(0.018)(0.003)
“Not recommended”−0.044−0.0500.006
(0.049)(0.050)(0.009)
Semi-elasticities (crisis)
“Great”−0.072a−0.050b−0.022a−0.098b−0.092b−0.006
(0.027)(0.025)(0.007)(0.045)(0.044)(0.011)
“Outstanding”−0.051a−0.035c−0.016a−0.072b−0.067b−0.005
(0.030)(0.018)(0.004)(0.030)(0.029)(0.007)
“Very good”−0.033b−0.021c−0.011a−0.049a−0.045b−0.004
(0.013)(0.013)(0.002)(0.018)(0.018)(0.004)
“Good”−0.014c−0.008−0.006a−0.026c−0.023c−0.004
(0.008)(0.008)(0.002)(0.014)(0.013)(0.002)
“Mediocre”0.0040.006−0.001−0.003−0.001−0.003
(0.008)(0.008)(0.003)(0.021)(0.022)(0.005)
“Not recommended”0.060b0.046c0.0140.0660.066−0.001
(0.026)(0.024)(0.009)(0.060)(0.063)(0.015)
R-squared0.5030.4960.5080.5920.5890.600
Observations8,0518,0518,0518,0518,0518,051
Notes: Firm-destination-time, grape, type, province, and HS fixed effects are included in (1) to (3). Product fixed effects are further included in (4) to (6). Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.
Notes: Firm-destination-time, grape, type, province, and HS fixed effects are included in (1) to (3). Product fixed effects are further included in (4) to (6). Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.

To evaluate the marginal effects of quality, the lower part of the table reports, for the pre-crisis and crisis periods, the semi-elasticities of each dependent variable to quality, evaluated at the mid score of each quality bin defined in Table 1.48 In column (1), the semi-elasticity of nominal export growth before the crisis is positive and significant for “Good,” “Very good,” “Outstanding,” and “Great” wines only, and its magnitude rises with quality (from 0.015 for “Good” to 0.061 for “Great” wines). During the crisis, the semi-elasticity is negative and significant for the higher quality wines only, and its magnitude again increases with quality (from -0.014 for “Good” to -0.072 for “Great” wines), while the positive coefficient for “Not recommended” wines indicates that their exports have grown.

Columns (2) and (3) show that the stronger growth in nominal exports for higher quality wines before the crisis was driven by changes in quantities rather than prices. In contrast, during the crisis, the lower growth for “Very good,” “Outstanding,” and “Great” wines was driven by lower growth in both quantities and prices (and to lower inflation only for “Good” wines). Columns (4) to (6) further include product fixed effects, and the results remain consistent (for prices the semi-elasticities are insignificant).

Using the estimates of columns (1) to (3), and separately for the two periods, Figure 4 plots the semi-elasticities of the growth in export values, volumes, and unit values to quality, evaluated at each quality score in the sample (i.e., from 68 to 96), and the corresponding 95 and 90 percent confidence intervals. Before the crisis, the semi-elasticities for export values and volumes are positive and significant for quality levels above 81, and are insignificant for unit values. During the crisis, the semi-elasticities are positive at low quality levels (between 68 and 71), and become negative for all wines above 81 for export values, and 86 for volumes. For unit values, the semi-elasticities are negative and significant for the wines above 79.

Figure 4.Export values, volumes, and unit values growth semi-elasticities to quality, by quality score and during the pre-crisis and crisis periods

B. Extensive Margin

By relying on first difference specifications, our analysis has so far explained changes at the intensive margin. As noted earlier, the changes we observe at the extensive margin are not negligible, however. Crozet and others (2012) show 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. To incorporate the extensive margin in our analysis, we proceed in several ways. First, we use as a dependent variable in equation (3) the mid-point export growth rate gijk,t of each wine k exported by firm i to country j in period t

where Xijk,t can take on zero values. This measure, which is symmetric around zero and is bounded between -2 and +2, has become standard in the labor economics literature as it allows to simultaneously study entries, exits, and continuing flows in an integrated framework (Bricongne and others, 2012; Davis, Haltiwanger, and Schuh, 1996).49 Second, we use as a dependent variable the first difference of ln (1+Xijk,t). For high levels of trade flows, ln (1+Xijk,t) ≃ ln Xijk,t, and for Xijk,t = 0, ln (1+Xijk,t) = 0.

We also estimate specifications in levels. As a benchmark, we first estimate equation (3), but using ln Xijk,t as a dependent variable. This approach excludes the zero trade observations, but includes the firm-product-destination triplets which exports are not observed between two consecutive periods. Second, we follow Crozet and others (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. Finally, we regress Xijk,t by Poisson Pseudo-Maximum Likelihood (Santos Silva and Tenreyro, 2006). Another way to address sample selection is to implement Heckman’s correction, as in Helpman, Melitz, and Rubinstein (2008), but this requires a variable that determines a firm’s destination-specific fixed costs of exporting, but not its variable trade costs, which is unavailable.

The results for nominal and real exports are reported in Table 17. Despite some variation in the magnitude of the effect of quality on trade across specifications, the resulting patterns are broadly supportive of our baseline results (the interaction is insignificant for export volumes in column 3). The level specifications in columns (3) to (5) indicate that the collapse in higher quality exports was characterized not only by slower growth, but also by less exports. Also, notice that compared to column (3), the interactions between quality and the crisis dummy in columns (4) and (5) are smaller once the extensive margin is accounted for.

Table 17.Extensive Margin
(1)(2)(3)(4)(5)
Panel A: Export values
quality×Dcrisis−0.024b−0.086b−0.037c−0.008a−0.019a
(0.009)(0.036)(0.019)(0.001)(0.001)
R-squared (or pseudo)0.6500.6160.7770.183
Panel B: Export volumes
quality×Dcrisis−0.023b−0.063b−0.031−0.008a−0.016a
(0.009)(0.030)(0.020)(0.001)(0.001)
R-squared (or pseudo)0.6520.6110.7870.185
Dependent variableMid-pointLog differenceLog levelsLog levelsLevels
EstimationOLSOLSOLSTobitPoisson
Zero observations includedYesYesNoYesYes
Observations12,42013,46814,11226,22426,224
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.

C. Post-Crisis Recovery

We argue that the crisis has led to a sharper contraction in higher than in lower quality exports. One way to validate our interpretation is to extend the analysis post-crisis, and to investigate whether the exports of higher quality goods grew more strongly once the world economy started to recover from the crisis. This exercise also helps us to establish whether the trade effects of the crisis were only temporary (Baldwin, 2009).50,51

Unfortunately, due to a glitch in the data collection, the customs dataset for 2010 reports the wine names as missing (while the firm names, export destinations, grapes, types, and vintage years are available). As a result, we cannot identify any of the wines exported during that year.52 Given the constraint to measure export growth between the same quarters to avoid the effects of seasonality, the earliest post-crisis data that can be used are observed for the 2011Q4–2012Q3 period. For each firm-product-destination triplet, the growth in export values, volumes, and unit values for the post-crisis sample is therefore measured between the 2008Q4–2009Q3 and the 2011Q4–2012Q3 four-quarter periods (i.e., with a two-year gap).

Our full sample now spans three different time periods as trade growth is computed for the pre-crisis, crisis, and post-crisis periods, respectively. For export values, we then estimate

where Dpost is a dummy variable equal to one for the 2011Q4–2012Q3 post-crisis period. The coefficient α1 captures the effect of quality on export growth before the crisis, while α2 and α3 measure the differential effects of quality on export growth during and after the crisis, respectively. For the pre-crisis and crisis periods, we have already shown that α1 > 0, α2 < 0, and α1 + α2 < 0. For the recovery, the effect of quality on export growth is given by α1 + α3. If export growth recovered more strongly for higher than for lower quality wines, as we would expect, α3 should either be insignificant, positive, or negative but with |α1| > |α3| such that α1 + α3 > 0 (i.e., the effect of quality on export growth during the recovery was positive, and its magnitude was either the same, larger, or smaller compared to before the crisis, respectively).

The results of estimating equation (5) for export values are reported in column (1) of Panel A in Table 18. As expected, export growth was stronger for higher quality wines both before and after the crisis, and fell to a larger extent during the downturn. As in Table 8, a one unit increase in quality strenghtened export growth by about two percentage points before the crisis, and subsequently lowered it by two percentage points. The positive coefficient on quality interacted with the post-crisis dummy indicates that a one unit increase in quality raised export growth by 6.3 percentage points (0.018+0.045) over the period. At first glance, this seems to suggest that the growth in higher quality exports post-recession has outpaced its pre-crisis performance. However, due to the unavailability of the 2010 customs data, recall that the growth rate post-crisis is calculated over a three-year period.

Table 18.Post-Crisis Recovery
(1)(2)(3)(4)(5)(6)(7)(8)
Panel A: Export values
quality0.018b0.031a−0.018c0.017c
(0.008)(0.008)(0.010)(0.009)
quality×Dcrisis−0.036a−0.049a−0.031b−0.042a
(0.013)(0.013)(0.014)(0.013)
quality×Dpost0.045a0.081a0.046a0.039b0.057a0.045b
(0.017)(0.018)(0.017)(0.019)(0.020)(0.021)
R-squared0.4910.4900.5050.4690.5770.5760.6370.609
Panel B: Export volumes
quality0.020b0.029a−0.0100.019b
(0.008)(0.007)(0.009)(0.009)
quality×Dcrisis−0.031b−0.040a−0.026b−0.034a
(0.013)(0.012)(0.013)(0.012)
quality×Dpost0.029b0.060a0.030b0.0270.042b0.034c
(0.014)(0.015)(0.014)(0.017)(0.017)(0.019)
R-squared0.4850.4850.5070.4570.5720.5720.6390.598
Panel C: Unit values
quality−0.0020.002−0.008a−0.002
(0.002)(0.003)(0.002)(0.002)
quality×Dcrisis−0.005−0.010a−0.004c−0.008b
(0.003)(0.003)(0.003)(0.003)
quality×Dpost0.016b0.021a0.015b0.012b0.016b0.011c
(0.007)(0.007)(0.007)(0.006)(0.007)(0.006)
R-squared0.4900.4890.5040.4580.5810.5810.6260.623
Product FENoNoNoNoYesYesYesYes
Sample included
Pre-crisisYesYesNoYesYesYesNoYes
CrisisYesYesYesNoYesYesYesNo
Post-crisisYesYesYesYesYesYesYesYes
Observations9,6179,6175,2735,9089,6179,6175,2735,908
Notes: Firm-destination-time, grape, type, province, and HS fixed effects are included in (1) to (4). Product fixed effects are further included in (5) to (8). Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.
Notes: Firm-destination-time, grape, type, province, and HS fixed effects are included in (1) to (4). Product fixed effects are further included in (5) to (8). Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.

We estimate variants of equation (5). Column (2) excludes the post-crisis interaction. Higher quality exports increased by more in “normal” periods (i.e., before and after the crisis combined), but fell to a larger extent during the crisis. Columns (3) and (4) exclude the pre-crisis and crisis data from the sample, respectively, and the results remain consistent. Finally, columns (5) to (8) report the same specifications but with product fixed effects included. The crisis and post-crisis interactions all remain negative and positive, respectively. The results for export volumes in Panel B are highly comparable, both quantitatively and qualitatively. In Panel C, there is some indication that inflation fell to a larger extent for higher quality wines during the crisis, but this finding is not always significant. In contrast, the inflation of higher quality wine exports was stronger after the downturn.

To conclude, our findings show that the crisis had heterogeneous effects on products differentiated by quality. The flight from quality in traded goods was only temporary, however, as evidenced by the recovery in higher quality exports that we observe after the end of the crisis period.

V. Economic Significance of Quality

We assess the contribution of quality to the evolution of wine exports. More precisely, we use our regression estimates to evaluate how Argentinean wine exports would have performed during the crisis under two alternative scenarios. First, we assume that the quality of all exported wines had increased at the highest level observed in our dataset (“quality upgrading”). Second, we consider the opposite extreme, and assume that the ratings of all wines had instead fallen at the lowest level of quality in the sample (“quality downgrading”). These two counterfactuals provide us with upper and lower bound estimates of the hypothetical performance of trade during the crisis due to changes in the quality composition of exports.53

We use our point estimates from regressing equation (3), reported in column (4) of Panel A in Table 8. Using the actual Wine Spectator scores, we derive the predicted values of export growth for each wine shipped to each destination country, and compare them against the predicted values obtained under the assumption that the quality ratings had all increased during the crisis at the highest level observed in the sample (which is equal to 96). On the basis of predicted export levels, we then find that Argentinean exports would have dropped by 38.94 percent, compared to an actual fall of 36.53 percent. In other words, quality upgrading during the crisis would have reduced nominal export growth by 2.4 additional percentage points. The opposite exercise, which assumes that the quality scores had fallen during the crisis at the lowest level of quality in the sample (which is equal to 68), predicts that exports would have dropped by 30.00 percent. The two opposite scenarios thus predict a difference in export performance of about nine percentage points, which is not negligible.

Using the same methodology, and the point estimates reported in column (4) of Panel B in Table 8, we then investigate the hypothetical performance of real export growth during the crisis, under the assumption that the quality of all wines exported had been either upgraded or downgraded during the period. Compared to an actual decrease in export volumes of 38.17 percent, we find that export quantities would have decreased by 40.63 percent with quality upgrading (at a value of 96), and by 32.39 percent with quality downgrading (at a value of 68). The difference in export growth predicted by the two opposite scenarios is therefore equal to about eight percentage points.

VI. Robustness

This section considers alternative specifications to check the robustness of our findings. Overall, the patterns we find are broadly supportive of the paper’s main conclusions.

(a) EndogeneityAshenfelter and Quandt (1999), Hodgson (2008), and Quandt (2007), among others, argue that wine ratings are subjective and imperfect measures of quality. Measurement error in the quality scores can thus create an endogeneity bias when explaining exports and prices. To address endogeneity, we estimate equation (3), and use the Parker scores as an instrument for the Wine Spectator ratings (both interacted with the crisis dummy variable) under the assumption that their measurement errors are uncorrelated.54 Column (1) of Table 19 reports the results. For both export values and volumes, the instrumented coefficients increase in magnitude, but qualitatively, our results continue to hold (the interaction is insignificant for unit values). The Kleibergen-Paap F statistic (equal to 437, with a critical value equal to 16, Stock and Yogo, 2005) rejects the null of weak correlation between the instrument and the endogenous regressor, and the first-stage regression shows that the Parker and Wine Spectator ratings, interacted with the crisis dummy, are positively correlated (the estimated coefficient is equal to 0.703, and is significant at the one percent level).

Table 19.Endogeneity and Robustness on Quality
(1)(2)(3)(4)(5)(6)(7)
Panel A: Export values
quality0.010−0.001
(0.019)(0.004)
quality×Dcrisis−0.072b−0.042b−0.101c−0.029c−0.023c−0.216a0.001c
(0.032)(0.017)(0.058)(0.015)(0.014)(0.082)(0.001)
R-squared0.5710.6100.5920.6000.5790.6000.599
Panel B: Export volumes
quality0.0130.000
(0.018)(0.003)
quality×Dcrisis−0.069b−0.039b−0.079−0.030b−0.019−0.180b0.001
(0.031)(0.017)(0.055)(0.014)(0.014)(0.077)(0.001)
R-squared0.5700.6050.5880.5950.5710.5940.594
Panel C: Unit values
quality−0.0030.000
(0.005)(0.001)
quality×Dcrisis−0.003−0.003−0.021c0.001−0.0030.000
(0.007)(0.004)(0.012)(0.003)(0.003)(0.000)
R-squared0.5710.6130.6000.5980.6070.605
QualityWSWeighted WSWS [1,6]Mean WSExcl. USUnit valueRanks
EstimatorIVOLSOLSOLSOLSOLSOLS
Observations4,3936,0638,0518,3917,18110,86910,869
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.
Notes: Firm-destination-time and product fixed effects are included. Robust standard errors adjusted for clustering by destination-time between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.

(b) Quality We compute a weighted average of the quality scores using the share of exports of each wine in the total exports of all wines with the same name, grape, and type, by destination and time period, as weights. This reduces our sample size as it restricts the analysis to the wines which original quality ratings are available. The correlation between weighted and unweighted quality is equal to 89 percent. The results are reported in column (2) of Table 19.

In column (3), to minimize possible noise in the measurement of quality when defined on a (50,100) scale, we use a variable which takes on values between one and six, where each value corresponds to one of the bins of the Wine Spectator (Table 1), and a larger value indicates a higher quality. In column (4), to include some unrated wines in the sample, we calculate a mean Wine Spectator rating by wine name and type, and assign this rating to all wines with the same name and type. Besides, as the Wine Spectator and Parker are both US-based ratings, one concern is that they may not capture taste preferences for quality in other destination countries. In column (5), we therefore drop the US from the sample. In most cases, our results continue to hold.

We also check if our results remain robust to measuring quality using unit values. Although this approach has been widely criticized (e.g., Khandelwal, 2010), its main advantage is to increase data coverage as all unrated wines can be included in the sample. First, we use as a proxy for quality the log mean unit value of exports across destinations and over time. Second, we rank for each exporter in each time period its wines by mean unit values in decreasing order (Mayer, Melitz, and Ottaviano, 2014). The wine with the highest unit value has a rank equal to one, the second a rank equal to two, etc., and we use these ranks as an inverted indicator of quality. Columns (6) and (7) report the results for unit values and product ranks, respectively. The results are in line with expectations.

(c) Sampling and Crisis Period In column (1) of Table 20, we extend the pre-crisis sample by one more year (up to 2005Q4). The pre-treatment sample therefore also includes trade growth computed between the 2005Q4–2006Q3 and the 2006Q4–2007Q3 four-quarter periods. In column (2), we further extend the pre-crisis sample up to 2003Q4 (we do not include the year 2002 as Argentina was in a recession). The pre-crisis sample therefore includes growth rates computed over four different periods.

Table 20.Other Robustness Checks
(1)(2)(3)(4)(5)(6)(7)(8)
Panel A: Export values
lagged dep. variable−0.389a
(0.033)
quality×Dcrisis−0.032a−0.025b−0.055a−0.046a−0.127b−0.027b−0.024−0.033b
(0.011)(0.011)(0.020)(0.018)(0.053)(0.011)(0.015)(0.016)
R-squared0.5610.5440.5900.5750.6850.6430.3080.660
Panel B: Export volumes
lagged dep. variable−0.410a
(0.035)
quality×Dcrisis−0.024b−0.018c−0.053a−0.044a−0.134b−0.019c−0.028c−0.028c
(0.011)(0.010)(0.017)(0.016)(0.052)(0.012)(0.016)(0.015)
R-squared0.5540.5330.5850.5710.6870.6510.3090.654
Panel C: Unit values
lagged dep. variable−0.340a
(0.050)
quality×Dcrisis−0.007a−0.007a−0.002−0.0010.007−0.008a0.005−0.006
(0.002)(0.002)(0.006)(0.005)(0.012)(0.003)(0.004)(0.004)
R-squared0.5640.5420.6100.5960.7000.6480.3340.636
Sample20052003Cr 09Q13Cr 09Q14VintageFullNo dest< 4.5l
Observations12,38118,6656,2287,4723,3335,6492,3358,641
Notes: Firm-destination-time and product fixed effects are included in (1)-(6) and (8). In (7), the firm-destination-time fixed effects are replaced by firm-time dummy variables. Robust standard errors adjusted for clustering by destination-time (and by firm-time in column 7) between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.
Notes: Firm-destination-time and product fixed effects are included in (1)-(6) and (8). In (7), the firm-destination-time fixed effects are replaced by firm-time dummy variables. Robust standard errors adjusted for clustering by destination-time (and by firm-time in column 7) between parentheses. a, b, and c indicate significance at the one, five, and ten percent levels.

We also vary the length of the crisis episode. In column (3), we drop the fourth quarter of 2008 from the crisis period. To avoid the effects of seasonality, the growth rates are again measured between the same quarters (in this case, over three-quarter periods). Instead, column (4) includes the fourth quarter of 2009 in the crisis period.

(d) Vintage Year As wine is an exhaustible resource, we have defined a wine product ignoring its vintage year. To ensure that this choice is not driving our results, we construct a new sample and define a product according to the name of the wine, its grape, type, and vintage year (the sample includes 99 firms exporting 1,025 different wines to 73 destination countries). The original Wine Spectator scores can therefore be used, but the highly unbalanced nature of this more disaggregated data reduces our sample size by more than half. Column (5) of Table 20 reports the results of estimating equation (3). For nominal and real exports, the interactions between quality and the crisis indicator are larger in magnitude compared to the baseline specifications of Table 8 (the results for unit values are insignificant), but most importantly, our results remain qualitatively similar.

(e) Mean Reversion The exports of higher quality wines grew the most before the crisis, and experienced the largest drops during the recession. To ensure that our findings are not just capturing mean reversion, we include lagged dependent variables. The results, reported in column (6) of Table 20, show that the effect of quality on trade during the crisis mattered beyond mean reversion.55,56

(f) Total Firm Exports Our analysis aims to identify the variation in export growth across products and over time. For our purposes, the destination country dimension is therefore not crucial. We aggregate our data to eliminate the destination country dimension, and calculate the exports of each product by each firm in each time period to all destinations. We estimate equation (3), and replace the firm-destination-time fixed effects by firm-time dummy variables, and cluster standard errors by firm-time. The results are reported in column (7) of Table 20.

(g) Small Volumes The volumes less than 4.5 liters are included in column (8) of Table 20. The effect of quality is slightly larger, but our results remain robust.

VII. Concluding Remarks

Using a unique dataset of Argentinean firm-level destination-specific export values and volumes of highly disaggregated wine products combined with experts wine ratings to measure quality, this paper shows that the global financial crisis has led to a flight from quality in traded goods. While empirical studies using different datasets and methodologies (Bems and di Giovanni, 2015; Berthou and Emlinger, 2010; Esposito and Vicarelli, 2011) reach conclusions that are complementary to ours, our analysis is the first to provide such evidence using firm-level data on traded goods.

As in any empirical work, our analysis suffers from a number of caveats. First, as we only observe Argentinean wine exports, we are unable to determine whether consumers in crisis-hit countries substituted from more expensive French or Italian wines towards cheaper Argentinean varieties.57 Indeed, Gil-Fournier (2008) notes that “many consumers will not want to pay abusive prices for European, mainly French, wines and they will probably decide to taste wines from other countries with a better relation between price and quality.” Second, we do not know whether our results remain relevant in a more general setting, and in particular whether they extend to other industries.

One promising avenue for future research would be to investigate the behavior of markups during the recent downturn, and in particular whether the decrease in inflation that we observe for the higher quality wines resulted from firms compressing their margins in order to preserve, or restore, export market shares.58 Indeed, 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 (see, for instance, Burstein and Jaimovich, 2012).

References

    AbiadA.MishraP.TopalovaP.2011. How does trade evolve in the aftermath of financial crises?IMF Working Paper WP/11/3.

    AhnJ.AmitiM.WeinsteinD.E.2011. Trade finance and the great trade collapse. American Economic Review: Papers and Proceedings101(3) 298302.

    • Search Google Scholar
    • Export Citation

    AlchianA.A.AllenW.R.1964. University Economics. Wadsworth PublishingBelmont, CA.

    AlessandriaG.KaboskiJ.P.MidriganV.2010. The great trade collapse of 2008–09: An inventory adjustment? IMF Economic Review58(2) 254294.

    • Search Google Scholar
    • Export Citation

    AmitiM.WeinsteinD.E.2011. Exports and financial shocks.Quarterly Journal of Economics126(4) 18411877.

    AndersonK.WittwerG.2013. Modeling global wine markets to 2018: Exchange rates, taste changes, and China’s import growth. Journal of Wine Economics8(2) 131158.

    • Search Google Scholar
    • Export Citation

    AndersonK.WittwerG.2015. Asia’s evolving role in global wine markets. CEPR Discussion Paper 10552.

    AshenfelterO.QuandtR.1999. Analyzing a wine tasting statistically. Chance12(3) 1620.

    BaldwinR.2009. The great trade collapse: Causes, consequences and prospects. VoxEU.org.

    BaldwinR.E.HarriganJ.2011. Zeros, quality, and space: Trade theory and trade evidence. American Economic Journal: Microeconomics3(2) 6088.

    • Search Google Scholar
    • Export Citation

    BehrensK.CorcosG.MionG.2013. Trade crisis? What trade crisis?Review of Economics and Statistics95(2) 702709.

    BemsR.di GiovanniJ.2015. Income-induced expenditure switching. Universitat Pompeu Fabramimeo.

    BemsR.JohnsonR.C.YiK.-M.2010. Demand spillovers and the collapse of trade in the global recession. IMF Economic Review58(2) 295326.

    • Search Google Scholar
    • Export Citation

    BemsR.JohnsonR.C.YiK.-M.2011. Vertical linkages and the collapse of global trade. American Economic Review: Papers and Proceedings101(3) 308312.

    • Search Google Scholar
    • Export Citation

    BermanN.MartinP.2012. The vulnerability of Sub-Saharan Africa to financial crises: The case of trade. IMF Economic Review60(3) 329364.

    • Search Google Scholar
    • Export Citation

    BernardA.B.JensenJ.B.ReddingS.J.SchottP.K.2009. The margins of US trade. American Economic Review: Papers and Proceedings99(2) 487493.

    • Search Google Scholar
    • Export Citation

    BerthouA.EmlingerC.2010. Crises and the collapse of world trade: The shift to lower quality. CEPII Working Paper 2010-07.

    BilsM.KlenowP.J.2001. Quantifying quality growth. American Economic Review91(4) 10061030.

    BricongneJ.-C.FontagnéL.GaulierG.TaglioniD.VicardV.2012. Firms and the global crisis: French exports in the turmoil. Journal of International Economics87(1) 134146.

    • Search Google Scholar
    • Export Citation

    BursteinA.EichenbaumM.RebeloS.2005. Large devaluations and the real exchange rate. Journal of Political Economy113(4) 742784.

    • Search Google Scholar
    • Export Citation

    BursteinA.JaimovichN.2012. Understanding movements in aggregate and product-level real exchange rates. University of CaliforniaLos Angeles, mimeo.

    • Search Google Scholar
    • Export Citation

    CastagninoT.D’AmatoL.SangiácomoM.2013. How do firms in Argentina get financing to export?European Central Bank Working Paper Series 1601.

    • Search Google Scholar
    • Export Citation

    ChenN.JuvenalL.2015. Quality, trade, and exchange rate pass-through. University of Warwickmimeo.

    ChoiY.C.HummelsD.XiangC.2009. Explaining import quality: The role of the income distribution. Journal of International Economics78(2) 293303.

    • Search Google Scholar
    • Export Citation

    ChorD.ManovaK.2012. Off the cliff and back: Credit conditions and international trade during the global financial crisis. Journal of International Economics87(1) 117133.

    • Search Google Scholar
    • Export Citation

    CoibionO.GorodnichenkoY.HongG.H.2015. The cyclicality of sales, regular and effective prices: Business cycle and policy implications. American Economic Review105(3) 9931029.

    • Search Google Scholar
    • Export Citation

    CrozetM.HeadK.MayerT.2012. Quality sorting and trade: Firm-level evidence for French wine. Review of Economic Studies79(2) 609644.

    • Search Google Scholar
    • Export Citation

    DavisS.J.HaltiwangerJ.C.SchuhS.1996. Job creation and destruction. The MIT PressCambridge, MA.

    EspositoP.VicarelliC.2011. Explaining the performance of Italian exports during the crisis: (Medium) quality matters. LLEE Working Paper 95.

    • Search Google Scholar
    • Export Citation

    EatonJ.KortumS.NeimanB.RomalisJ.2015. Trade and the global recession. Penn State Universitymimeo.

    EngelC.WangJ.2011. International trade in durable goods: Understanding volatility, cyclicality, and elasticities. Journal of International Economics83(1) 3752.

    • Search Google Scholar
    • Export Citation

    FajgelbaumP.GrossmanG.HelpmanE.2011. Income distribution, product quality, and international trade. Journal of Political Economy119(4) 721765.

    • Search Google Scholar
    • Export Citation

    FeenstraR.C.RomalisJ.2014. International prices and endogenous quality. Quarterly Journal of Economics129(2) 477527.

    FreundC.2009. The trade response to global downturns – Historical evidence. World Bank Policy Research Working Paper 5015.

    Gil-FournierJ.M.2008. Argentina looks for silver lining. Available at http://www.ofournier.com.

    GopinathG.ItskhokiO.NeimanB.2012. Trade prices and the global trade collapse of 2008–09. IMF Economic Review60(3) 303328.

    • Search Google Scholar
    • Export Citation

    GriffithR.O’ConnellM.SmithK.2013. Food expenditure and nutritional quality over the great recession. Institute for Fiscal Studies Briefing Note BN143.

    • Search Google Scholar
    • Export Citation

    HaddadM.HarrisonA.HausmanC.2010. Decomposing the great trade collapse: Products, prices, and quantities in the 2008–2009 crisis. NBER Working Paper 16253.

    • Search Google Scholar
    • Export Citation

    HallakJ.C.2006. Product quality and the direction of trade. Journal of International Economics68(1) 238265.

    HelpmanE.MelitzM.J.RubinsteinY.2008. Estimating trade flows: Trading partners and trading volumes. Quarterly Journal of Economics123(2) 441487.

    • Search Google Scholar
    • Export Citation

    HodgsonR.T.2008. An examination of judge reliability at a major US wine competition. Journal of Wine Economics3(2) 105113.

    HoekmanB.2015. The global trade slowdown: A new normal?VoxEU.org.

    HummelsD.KlenowP.J.2005. The variety and quality of a nation’s exports. American Economic Review95(3) 704723.

    HummelsD.LeeK.Y.2012. Income elastic goods and the great trade collapse: Evidence from microdata. Purdue Universitymimeo.

    HummelsD.SkibaA.2004. Shipping the good apples out? An empirical confirmation of the Alchian-Allen conjecture. Journal of Political Economy112(6) 13841402.

    • Search Google Scholar
    • Export Citation

    IacovoneL.ZavackaV.2009. Banking crises and exports: Lessons from the past. World Bank Policy Research Working Paper 5016.

    JaimovichN.RebeloS.WongA.2015. Trading down and the business cycle. CEPR Discussion Paper 10807.

    JohnsonR.C.2012. Trade and prices with heterogeneous firms. Journal of International Economics86(1) 4356.

    KeeH.L.NeaguC.NicitaA.2013. Is protectionism on the rise? Assessing national trade policies during the crisis of 2008. Review of Economics and Statistics95(1) 342346.

    • Search Google Scholar
    • Export Citation

    KhandelwalA.2010. The long and short (of) quality ladders. Review of Economic Studies77(4) 14501476.

    KuglerM.VerhoogenE.2012. Prices, plant size, and product quality. Review of Economic Studies79(1) 307339.

    LevchenkoA.A.LewisL.T.TesarL.L.2010. The collapse of international trade during the 2008–09 crisis: In search of the smoking gun. IMF Economic Review58(2) 214253.

    • Search Google Scholar
    • Export Citation

    LevchenkoA.A.LewisL.T.TesarL.L.2011. The “collapse in quality” hypothesis. American Economic Review: Papers and Proceedings101(3) 293297.

    • Search Google Scholar
    • Export Citation

    MaliziaG.2008. In our current situation there is no choice but lowering prices. Available at http://www.winesur.com.

    ManovaK.ZhangZ.2012a. Export prices across firms and destinations. Quarterly Journal of Economics127(1) 379436.

    ManovaK.ZhangZ.2012b. Multi-product firms and product quality. NBER Working Paper 18637.

    MayerT.MelitzJ.M.OttavianoG.I.P.2014. Market size, competition, and the product mix of exporters. American Economic Review104(2) 495536.

    • Search Google Scholar
    • Export Citation

    NevoA.WongA.2015. The elasticity of substitution between time and market goods: Evidence from the great recession. NBER Working Paper 21318.

    • Search Google Scholar
    • Export Citation

    QuandtR.E.2007. On wine bullshit: Some new software?Journal of Wine Economics2(2) 129135.

    Santos SilvaJ.M.C.TenreyroS.2006. The log of gravity. Review of Economics and Statistics88(4) 641658.

    SchottP.K.2004. Across-product versus within-product specialization in international trade. Quarterly Journal of Economics119(2) 647678.

    • Search Google Scholar
    • Export Citation

    StockJ.H.YogoM.2005. Testing for weak instruments in linear IV regression in: AndrewsD.W.K.StockJ.H. (Eds) Identification and Inference for Econometric Models: Essays in Honor of Thomas Rothenberg. Cambridge University PressCambridge pp. 80108.

    • Search Google Scholar
    • Export Citation

    ThorntonJ.2013. American Wine Economics. An Exploration of the US Wine Industry. University of California PressBerkeley and Los Angeles, California.

    • Search Google Scholar
    • Export Citation

    TradC.2008. The crisis has hit at a moment when Argentinean wine is in its best shape. Available at www.winesur.com.

    VerhoogenE.2008. Trade, quality upgrading and wage inequality in the Mexican manufacturing sector. Quarterly Journal of Economics123(2) 489530.

    • Search Google Scholar
    • Export Citation

    VesethM.2008. How will the economic crisis affect wine?Available at http://wineeconomist.com.

    World Trade Monitor2014. CPB Netherlands Bureau for Economic Policy Analysis.

1

For evidence that the income elasticity of demand increases with quality, see Bils and Klenow (2001) who estimate “quality Engel curves,” and show that a higher income increases the consumption of higher quality goods. Berthou and Emlinger (2010) and Esposito and Vicarelli (2011) find that the income elasticity of imports rises with quality. Also see Choi, Hummels, and Xiang (2009).

2

Burstein, Eichenbaum, and Rebelo (2005) and Jaimovich, Rebelo, and Wong (2015) define a “flight from quality,” or a “trading down in quality,” as a substitution towards lower quality goods in response to contractionary devaluations or to recessions, respectively. Another example is the civil war in Syria which has reduced tourism in neighboring Lebanon where, as a result, the imports of luxury goods collapsed (see “Lebanon suffers flight from quality as Syria war hits economy,” The Financial Times, October 4, 2013).

3

Recall that with homothetic preferences, all goods have unitary income elasticities.

4

It is unlikely that all households respond homogeneously to aggregate income shocks. For instance, during the crisis, some households may have consumed goods in smaller quantities, but with a higher quality content. Ultimately, the effect of the crisis on the total consumption of higher quality goods depends on the aggregation over households.

5

As explained below, we however define a “product” according to the name of the wine, its type, and grape, but ignore the vintage year. As wine is an exhaustible resource, we want to avoid the possibility that a wine is exported less or drops out from the sample during the crisis only because its supply is running out.

6

Evidence suggests that the fine wine market devoted to investors and collectors has remained immune to the global financial crisis. Argentinean wines are not, however, traded by wine auction houses such as Christies or Sothebys, which predominantly specialize in wines from the Bordeaux region as they are the most popular among investors and collectors.

7

Bems and di Giovanni (2015) use similarly disaggregated product-level data but at the retail-level.

8

Levchenko, Lewis, and Tesar (2011) “caution against using unit values as proxies for quality, especially when drawing conclusions about the role of quality differentiation in the recent trade collapse.”

9

As the production of higher quality goods is more labor intensive than that of lower quality goods, Jaimovich and others (2015) claim that trading down during recessions reduces labor demand, and therefore employment.

10

Amiti and Weinstein (2011) stress that the key question of the trade collapse literature is not why trade fell, but rather why trade has been much more disrupted than domestic output. Our paper is concerned with how the crisis has affected the composition of trade, and does not look at its impact on domestic production.

13

This approach is criticized by, among others, Khandelwal (2010).

15

As we explain later, we also have data for the 2010 to 2012 period. However, due to a glitch in the data collection, the wine names cannot be identified in the 2010 sample.

16

Wholesalers and retailers represent a very small share of the sample. Once the customs data are merged with the quality ratings, these firms all drop out from the sample.

17

Due to the highly disaggregated nature of the data, aggregating at a monthly frequency would significantly reduce the number of log changes in exports and prices that can be measured.

18

This assumption sounds reasonable as evidence suggests that the quality of Argentinean wines does not vary much across vintage years. If wine producers conclude that the grapes grown during a particular year do not satisfy their quality standards, they may decide not to use them in order to preserve the reputation of a wine. In addition, modern winemaking techniques allow to minimize the variation in wine quality over time (Thornton, 2013). As we show later, our results remain robust to letting wine products vary across vintage years.

19

For instance, Chor and Manova (2012) define August 2009 as the end of the crisis.

20

The growth rates are calculated year-on-year in order to minimize seasonal effects. Quarterly total exports are from the International Financial Statistics of the International Monetary Fund.

21

Castagnino, D’Amato, and Sangiácomo (2013) observe that the largest Argentinean firms tend to export continuously, while small and medium-sized firms are instead sporadic exporters. Summing exports over four quarters allows us to increase the sample coverage of the small and medium-sized firms.

22

As we show later, our results remain robust to alternative starting or ending dates.

23

As pointed out by Levchenko, Lewis, and Tesar (2010), an alternative would be to measure the percentage drop in trade from the peak to the trough of the recession, but this measure would be contaminated by seasonality.

24

Subjective quality refers to the evaluation of consumers based on personal tastes and preferences, whereas objective quality rates the characteristics of a wine which are unrelated to personal preferences such as balance, complexity, finish, concentration, or flavor intensity. The role of experts is to provide an objective assessment of quality (Thornton, 2013).

25

In the figure we report in Chen and Juvenal (2015), one difference is that a few wines are rated as “Not recommended” (a value of one) by the Wine Spectator, and as “Outstanding” (a value of five) by Parker.

26

The eight countries dropping during the crisis are Antigua and Barbuda, Bahamas, Barbados, Belize, Croatia, Cyprus, Hungary, and Suriname, while the three entering are Grenada, Indonesia, and Sri Lanka.

27

Although our findings show that the intensive margin dominates, its contribution is smaller than in Behrens and others (2013) where the intensive margin explains 97 percent of trade growth.

28

Considering both margins, the total change in wine exports is equal to (0.9467×0.9316×0.9829×0.8135−1) × 100 = −29.48 percent. If we further decompose the intensive margin into changes in mean quantities and prices, the total change in wine exports is given by (0.9467×0.9316×0.9829×0.7801×1.0428−1)×100 = −29.48 percent.

29

Our fixed effects control for the various determinants of the trade collapse that have been studied in the literature.

30

“Old World” wines are labelled by the region where they are produced and where the grapes are grown, whereas “New World” wines are classified by their grapes which can be grown in different provinces. This is the reason why we control for the province of origin of the grapes (Chen and Juvenal, 2015).

31

As the mean Wine Spectator and Parker quality scores remained the same before and during the crisis, we can rule out that our results are driven by a lower quality reducing the demand for wine exports during the recession.

32

Changes in exchange rates also have heterogeneous effects on the exports and the prices of goods differentiated by quality. In Chen and Juvenal (2015) we show, both theoretically and empirically, that a real depreciation increases export prices by more, and export volumes by less, for higher quality goods. We estimated equation (3), including changes in bilateral exchange rates interacted with quality, but the estimated coefficients were insignificant, most likely because our sample does not provide sufficient time-series variation to uncover significant exchange rate effects.

33

The data on real GDPs per capita (in US dollars) are from the Penn World Tables, while household real final consumption expenditures per capita (in US dollars) and the unemployment rates are from the World Bank’s World Development Indicators. All variables are measured annually for the years 2007-2009. One shortcoming of using annual data is that in many countries, aggregate demand was strong both at the start of 2007 and at the end of 2009. Quarterly data are however not available for most emerging markets and developing economies.

34

Behrens and others (2013), Bems and others (2010), and Freund (2009) find that the sensitivity of trade to GDP growth increased during the crisis. Instead, we look at the effect of changes in aggregate demand on the semi-elasticity of trade to quality. Including a triple interaction between changes in aggregate demand, quality, and the crisis dummy yields insignificant results.

35

Chor and Manova (2012) observe that, by the end of 2008, the fall in US imports was larger than in US exports, reflecting “the particular sharp decline in consumer sentiment and import demand in the US relative to other countries.”

36

Behrens and others (2013) observe that Belgium’s trade with China was much less affected during the crisis.

37

Compared to Western countries, China’s per capita wine consumption is low. However, due to the size of its adult population and the fact that “income growth has led to a burgeoning middle class and enriched their elite” (Anderson and Wittwer, 2013), China’s demand for wine has soared. China has expanded its wine production dramatically, and is today among the world’s five largest wine producers, but as “supply expansion has not been able to keep up with China’s growth in demand […], wine imports have surged” (Anderson and Wittwer, 2015).

39

The results (available upon request) remain similar if we also include interactions between quality and a dummy for each country group. Interacting quality with GDP per capita growth separately for each country is not informative as for each country, income growth only varies over time. Similarly, product-time fixed effects cannot be included.

40

In support of our findings, Kym Anderson has argued, at the American Association of Wine Economists meeting in 2011, that “the removal of wine import taxes in Hong Kong has led to a large increase in demand for iconic wines.”

41

The increase in patriotism during recessions can be instigated by policymakers. Examples include the call in January 2009 from the industry minister in Spain for the population to shop Spanish in order to fight the recession, or the inclusion of a “Buy American” provision in the US fiscal stimulus package that was enacted in February 2009.

42

Consumers could also substitute between wines imported from different countries. Due to data constraints, we are unable to address this possibility.

43

The results remain similar if we also include interactions between quality and a dummy for each group of countries.

44

The correlation between the five largest wine producing nations and the self-sufficient countries is equal to 0.49.

45

Using the 2005-2006 World Values Survey, we explored whether the extent to which some countries are more patriotic than others could also explain our results. Respondents were asked the question “How proud are you to be [nationality]?” For each country, we calculated the share of respondents who were either “Very proud” or “Quite proud” of their citizenship. Due to limited coverage, our sample was reduced to 36 countries. We interacted the share of patriotic respondents with quality, and included it as a regressor in equation (3), but its coefficient was insignificant.

46

Using data (which are not publicly available) on the domestic and foreign debt held by Argentinean firms, Castagnino and others (2013) show that small and medium-sized firms are more restricted in accessing external finance than large firms.

47

Domestic credit conditions in Argentina became tighter during the crisis, as evidenced by the daily interbank lending rate that rose from an average of 8.8 to 10.9 percent between the 2007Q4–2008Q3 and 2008Q4-2009Q3 periods (data from Chor and Manova, 2012).

48

The semi-elasticities are evaluated at a value of 62 for “Not recommended,” 77 for “Mediocre,” 82 for “Good,” 87 for “Very good,” 92 for “Outstanding,” and 97.5 for “Great” wines.

49

The mid-point growth rate is equal to +2 for an entry, -2 for an exit, takes on values between zero and +2 if flows increased over time, and between -2 and zero if they decreased over time.

50

Anderson and Wittwer (2013) note that, “due to the fall in income and wealth,” the crisis has brought “a temporary decline in the quantity and quality of wine demanded in traditional markets.”

51

World trade growth experienced a sharp recovery in 2010, but slowed down markedly in the course of 2011 and has remained sluggish since then. Evidence suggests that a combination of cyclical and structural factors is negatively impeding global trade growth (Hoekman, 2015). Also, see Eaton, Kortum, Neiman, and Romalis (2015).

52

According to Nosis, the issue with the 2010 sample comes from the Argentinean customs, and the data are, unfortunately, unlikely to be updated in the future.

53

These predictions are derived from reduced-form specifications and therefore only provide a rough estimate of the contribution of quality to export growth.

54

In Chen and Juvenal (2015), we use weather and geography-based instruments that vary across vintage years.

55

Due to the short time dimension of our data, we cannot use dynamic panel data estimation techniques to deal with the well-known problem of regressing within-group equations with lagged dependent variables.

56

In results available upon request, we show that our findings for the post-crisis recovery are robust to mean reversion.

57

Levchenko and others (2011) investigate whether this type of cross-country substitution within a specific product category is present in the data, but find no significant evidence. Also, households may have switched from wine to cheaper alcoholic beverages, such as beer, but due to data constraints we are unable to explore this possibility.

58

Anecdotal evidence suggests that during the crisis, wine producers had “no choice but lowering prices” in order to offer “wines that consumers can afford” (Malizia, 2008). Also, see Thornton (2013).

Other Resources Citing This Publication