Geographical and industrial spillovers in entry decisions across export markets

Abstract

This article addresses sequential entry decisions in export markets. It focuses on externalities derived from previous export activity in countries close to those for which a potential entry decision is made (geographical spillovers) and externalities derived from previous presence of other firms in the same industry (industrial spillovers). The empirical analysis uses Spanish microdata for the period 2000–2010 in a firm decision model that also integrates country and industry characteristics. The results suggest that these two types of spillovers have a positive effect in explaining entry decisions in new export markets, though both are smaller in magnitude than the effects coming from previous presence in the same specific destination.

Keywords:

• sequential entry
• spillovers
• export activity
• firm heterogeneity

JEL Classification:

• F10
• F14

Notes

¹ Empirical evidence on the importance of multi-market and multi-product exporters on total export value is provided by Bernard et al. (2007) for the US, Mayer and Ottaviano (2008) for France and Bastos and Silva (2010) for Portugal, among others.

² Many studies of internationalization for Spanish firms use the Encuesta Sobre Estrategias Empresariales (ESEE). However, that database only provides quadrennial information on export destinations aggregated in four broad geographical areas.

³ Further information about the database can be found at http://directorio.camaras.org/.

⁴ The overall volume of firm exports is grouped in three segments: less than one hundred thousand euros, between that amount and one million euros, and more than one million euros.

⁵ Castillo-Giménez et al. (2011) analyse the determinants of a firm’s export decision by focusing on the influence of proximity to other exporters. Moreover, Esteve-Pérez et al. (2013) investigate the duration of Spanish firms’ trade relationships by applying a survival analysis.

⁶ We should remember that the product classification followed by the data set is highly aggregated, so this comparison should be taken with caution.

⁷ Levinsohn and Petrin (2003) propose a semiparametric model that uses intermediate inputs (e.g., materials and energy) as proxies for unobserved productivity. Using this approach, gross revenue, capital stock, number of employees and materials (for each firm/year) are used to estimate and construct this TFP measure.

⁸ Of course, product diversification is a strategy closely related to firm size. However, note that the effect of firm size is already controlled for in the empirical analysis.

⁹ We do not consider the existence of Free Trade Agreements (FTAs) in our country classification. The following example clarifies this issue. Only one of the three NAFTA countries (Mexico) is currently a signatory to a regional trade agreement with the EU. Thus, a Spanish firm that was exporting to Mexico in t – 1 does not have any incentive to enter another country that also belongs to NAFTA in period t, once non-FTA variables are controlled for.

¹⁰ See the Appendix for more details on the elaboration of both variables.

¹¹ In complementary regressions, size was measured with the three segments of overall volume of exports and results remain unchanged.

¹² The correlation between the variables related to TFP and firm size is quite high, which might suggest the lack of significance in the TFP when both variables are considered simultaneously. However, coefficients of these variables are clearly significant when they are included separately. For this reason, Column (iii) of Table 3 only includes the variable related to TFP.

¹³ To check the robustness of the results, we have also considered the classification of countries suggested by The World Bank. This classification distinguishes between seven large aggregate areas: East Asia and Pacific, Europe and Central Asia, North America, Latin America and the Caribbean, the Middle East and North Africa, South Asia, and Sub-Saharan Africa. The results also confirm the positive effect of the geographical spillovers on entry decisions.

¹⁴ As was previously mentioned, this analysis excludes those countries in which the number of occurrences (firms exporting to that country in a specific year) is lower than 20. To check the robustness of the results, we have repeated the analysis excluding those countries in which the number of occurrences is lower than 200. By considering this threshold, the number of countries is reduced to 151 and the significance of the considered variables remains unchanged.

Additional information

Funding

The authors gratefully acknowledge financial support from the Spanish Ministry of Science and Innovation [reference ECO2010-18947].