Understanding how domestic companies could be supported in their efforts to break into export markets and diversify their range of export products is an important concern for policymakers. Indeed, export diversification is viewed as a way of stimulating growth by policymakers and academics alike (Rivera-Batiz and Romer 1991, Grossman and Helpman 1991, Hausmann et al 2007).
An effective export-promotion policy cannot be designed without understanding the drivers of exporting at the micro level. Many empirical and theoretical contributions have demonstrated that exporters outperform other firms in terms of productivity and size (see Bernard and Jensen 1999, Clerides et al 1998, Melitz 2003 and the subsequent studies). But is it due to nature or nurture? Very little attention has been devoted to the key question of how firms can learn to export.
Exporters are made, not born
Building on the literature pointing out the importance of quality in explaining trade patterns, our recent study (Iacovone and Javorcik 2012) provides, for the first time, direct evidence on quality upgrading taking place in preparation for penetrating exports markets. Relying on very detailed information from Mexico, the study analyses dynamics at the plant-product level before and after entry into export markets during an export boom (see Figure 1). The unique data on product sales at home and abroad disaggregated by producer and year allow us to search for evidence of preparation for exporting and contribute to the debate on whether exporters are born or made through the lens of quality. The analysis focuses on the 1994–2004 period when Mexican producers supported by a sharp devaluation and the creation of NAFTA increased their efforts to penetrate the US market and introduced a large number of new export products.
Figure 1. Mexican export boom
Our study demonstrates that producers that export a particular product tend to enjoy a price premium on the domestic market relative to other companies producing the same product. This premium appears exactly one year before the product is introduced into export markets and is inexistent two and three years before exporting. There is no evidence of upgrading after the product’s entry into export markets.
Exploiting the introduction of NAFTA as a quasi-natural experiment, our study uses the announced schedule of tariff cuts taking place under NAFTA (agreed prior to the agreement coming into effect) as an instrument to identify the causal link between quality upgrading and entry into exports market. The results of the instrumental variable approach, using the anticipated tariff cuts under NAFTA and each producer’s initial market share in a given product, suggest the increase in price premium is a response to the anticipated improvement in the market access on the part of high performers. This finding is consistent with producers consciously improving their products in preparation for introducing them into export markets. It is in line with the theoretical predictions of Constantini and Melitz (2007) that the anticipation of future liberalisation induces producers to innovate ahead of liberalisation and thus also ahead of their anticipated, but yet unrealised, entry into export market.
Other supporting evidence
The view that producers consciously upgrade their products in preparation for introducing them abroad is also supported by anecdotal evidence. In 2007, we interviewed an executive from a leading Mexican company producing fruit and vegetable juices. When asked what it takes for a company like his to become an exporter, the executive pointed to “quality, quality, and quality.” According to the executive, the first dimension of quality relevant to exporting was bringing the product up to the level which satisfies foreign sanitary and phytosanitary standards, which tend to be higher in industrialised countries than in Mexico.
The second dimension of quality was making the product appeal to the tastes of foreign consumers. Consumers in the US (the major export market for this producer) demand higher-quality products than the average Mexican buyer. For instance, they prefer juices closer in taste to fresh juices than products from concentrates. Therefore, the company invested in a new technology to produce such juices. They were first sold domestically targeting higher-end Mexican consumers and subsequently they were introduced in the export market. The decision to introduce such juices was made with the export market in mind as the company recognised that the local market for such a high-end product is quite limited.
The third dimension of quality relevant to juice producers is packaging. While Mexican consumers prefer cartons, US buyers have a preference for plastic and glass containers. In the juice industry, package attractiveness plays a very important role. To improve the quality of its packaging, the company opted for a new technology where containers are covered with sleeves on which product labels are printed, as this produces a more attractive appearance than printing directly on a container.
Another piece of evidence supporting the story of quality upgrading comes from the Survey on Employment, Science, Technology and Training (ENESTyC) conducted in 1992 and 2001 which reports information on whether or not an establishment uses an automated quality control system. As illustrated in Figure 2, we find that new and established exporters were more likely to introduce an automated quality control process in sectors experiencing a large improvement in market access under NAFTA than in sectors where little improvement was seen. Among producers focused solely on the Mexican market, no such difference was registered between the two types of sectors.
Figure 2. Probability of introducing a quality-control system
Why would product upgrading take place one year prior to exporting?
Most likely, Mexican producers learn from potential foreign buyers that there is interest in their product, provided they meet quality requirements. To meet the buyer’s expectations, Mexican producers upgrade their product which may require additional investment in physical assets. This investment is worthwhile as its cost can be spread over the large export market.
This view is consistent with the evidence on producers’ investment behaviour presented in Figures 3 and 4. The first figure plots the distribution of real investment (in log) for (i) exporters that will introduce a new export variety in the next period and (ii) exporters that will not do so but will continue exporting. As visible from the figure, 77% of plants in the former group invest in physical assets, as opposed to 71% in the latter group (see the spike around zero). Moreover, among those investing, exporters introducing a new export variety in the following period tend to invest a larger amount. The differences in the investment pattern are even more pronounced when non-exporters are compared to producers that will start exporting the following year (see Figure 3). While 70% of future exporters invest in physical assets, this is true of only half of producers that will remain non-exporters next period.
Figure 3. Distributions of investment in capital, exporters introducing a new variety vs other exporters
Figure 4. Distributions of investment in capital, future exporters vs non-exporters
After the quality upgrading takes place, Mexican producers present the new version of the product to potential buyers. As it takes time to finalise a deal (further improvements may be needed, price negotiations may be lengthy, larger volume may be needed), producers start selling the product in Mexico. The described scenario is consistent with the observation that product upgrading would take place about a year prior to the variety being introduced in export markets.
The above findings suggest that helping prospective exporters improve product quality may be a vital part of export promotion activities.