Abstract
This paper investigates the relationship between innovation and the export behavior of firms using data from a representative panel of Spanish firms over 1991–2002. It presents a simple theoretical model of the firm decision to export and innovate that guides the econometric analysis. Consistent with the predictions of the theoretical model, the econometric results suggest a positive effect of firm innovation on the probability of participation in export markets. The results further reveal the heterogeneous effects of different types of innovations on the firm export participation. In particular, product upgrading appears to have a larger effect on the firm export participation than the introduction of cost-saving innovations. These findings are robust to firm unobserved heterogeneity, dynamic specifications, and to the use of instrumental variables to control for the potential endogeneity between innovation and exporting.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Evidence in favour of self-selection of better performing firms into export markets is provided by Clerides et al. (1998) for Colombia, Mexico and Morocco; Bernard and Jensen (1999) for the US; and Delgado et al. (2002) for Spain. Some papers find evidence of learning by exporting including van Biesebroek (2006) for Côte-d’Ivoire and de Loecker (2007) for Slovenia. For a review of this literature see Greenaway and Kneller (2007) and Wagner (2007).
The paper does not explore the effects of other sources of knowledge acquisition such as acquired technology or purchases of technology from abroad on exporting, which may arguably be also important for Spanish firms.
For the case of many varieties N the firm i′s market share s i is small and the elasticity of demand η is equal to the elasticity of substitution σ: η = σ − s i (σ − 1) because it is generally assumed that an individual firm i market share s i is small so the second term becomes negligible.
I would like to thank one of the referees for suggesting to consider different demand elasticities for the domestic and export market.
Figure 1 panel A shows that the data supports this assumption. The figure shows that the productivity distribution of innovating firms lies to the right of the productivity distribution of non-innovating firms, suggesting that innovators are more productive as assumed by the theoretical model.
The data does not support this case as we observe that about 60% of firms export compared to only about a third of them that innovate, which tends to suggest that the fixed costs to innovate are higher than the fixed costs to export. Therefore in the descriptive statistics that follow I will focus on the case where [(f I=1 − f I=0) > f x ] .
For more details on the survey design see Fariñas and Jaumandreu 1999.
The data used in this study has been previously used by Delgado et al. (2002) to study the productivity of exporting firms and Fariñas and Ruano (2005) to study firm heterogeneity and market selection. The data is cleaned with respect to unlikely values and large spikes in the variables used in the empirical analysis. The first year of the sample is lost because there is no data available for capital necessary to construct productivity. Due to this cleaning less than 2% of the observations are lost, so the sample remains highly representative.
The Olley and Pakes (1996) methodology controls for the endogeneity in input choice by using the firm investment decision to proxy for unobserved productivity shocks and for firm survival introducing a survival rule into the estimation method. The endogeneity of input choice and the selection bias are well-known to the literature and have been documented extensively. For a review of the main methodological issues see Ackerberg et al. (2007); van Beveren (2007); van Biesebroeck (2007) for a discussion of these biases.
The results are robust to alternative specifications where the explanatory variables are lagged two periods. These results are available from the author upon request.
By including only one lag of the export status variable, I assume that firms completly loose their sunk start-up costs if they are absent from market 1 year. This is in line with previous empirical evidence suggesting that sunk start-up costs depreciate very quickly and that firms that most recently exported 2 years ago have to pay nearly as much to re-renter foreign markets as first time exporters (Das et al. 2007; Máñez et al. 2008; Roberts and Tybout 1997).
Some of the known shortcomings of the linear probability model are: non-normal errors, non-constant error variance, non-linearity and non-sensical predictions that can create predicted values that are not bounded between zero and one. For further insights into the advantages of the probit model versus the linear probability model see Green (2008).
The definition of research and development (R&D) used by the ESEE survey is the standard international definition as given by the Frascati Manual. I verified that all the firms that answered yes to the question whether they conducted R&D activities in a given year also reported some positive R&D expenditures.
The process innovation variable takes value 1 if the firm says that it has introduced a process innovation (new machinery, new organization methods, or both) during the survey year. Following this definition, the introduction of a process innovation may be the result of imitation. Evidence by Huergo and Jaumandreu (2004) using the same data set however shows that firms tend not to report changes which are generalised in the market as innovative productive modifications.
This literature suggests that acess to larger markets increases the efficiency of firms as they can exploit scale economies. There are several papers that investigate the scale effects of opening-up to trade among others, for example, Trefler (2004) analyzes the effects of the Canada-US Free Trade Agreement on the efficiency of Canadian firms.
In the instrumental variables estimations I estimate linear models rather probit models because instrumental variables regressions (IV) within the probit framework require very strong assumptions (See Carrasco (2001) for details on IV estimation in the probit framework).
The variable motivated directly by the theoretical model is the R&D dummy for which the instrument of public R&D support is strongly correlated with. Although intuitive and significant in the first stage 2SLS (Table 7, columns 2, 3, and 4), the other three proxies for innovation should be interpreted with caution, because their correlation with the instrument is weaker than for the R&D dummy.
References
Ackerberg, D., Benkard, L., Berry, S., & Pakes, A. (2007). Econometric Tools for Analyzing Market Outcomes. In J.J. Heckman (Ed.), Handbook of Econometrics, Volume 6. Amsterdam: Elsevier.
Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. Review of Economic Studies, 58(2), 277–297.
Aw, B. Y., & Hwang, A. R. (1995). Productivity and the export market: A firm level analysis. Journal of Development Economics, 47(2), 313–332.
Aw, B. Y., Roberts, M. J., & Winston, T. (2007). Export market participation, investments in RD and worker training, and the evolution of firm productivity. The World Economy, 14(1), 83–104.
Aw, B. Y., Roberts, M. J., & Yi, D. Xu (2009). R&D investments, exporting and productivity dynamics. (NBER Working Paper 14670) Cambridge, MA.
Bartelsman, E., & Doms, M. (2000). Understanding productivity: Lessons from longitudinal microdata. Journal of Economic Literature, 38(3), 569–594.
Bernard, A. B., & Jensen, J. B. (1995). Exporters, jobs, and wages in US manufacturing: 1976–1987. Brooking Papers on Economic Activity: Microeconomics, 67–119.
Bernard, A. B., & Jensen, J. B. (1999). Exceptional exporter performance: Cause, effect, or both? Journal of International Economics, 47(1), 1–25.
Bernard, A. B., & Wagner, J. (2001). Export entry and exit by German firms. Weltwirtschaftliches Archiv/Review of World Economics, 137(1), 105–123.
Bernard, A. B., Eaton, J., Jensen, B. J., & Kortum, S. (2003). Plants and productivity in international trade. American Economic Review, 93(4), 1268–1290.
Bernard, A. B., & Jensen, B. J. (2004). Why some firms export? Review of Economics and Statistics, 86(2), 561–569.
Bernard, A. B. (2007). Firms in international trade. Journal of Economic Perspectives, 21(3), 105–130.
van Beveren, I. (2007). Total factor productivity estimation: A practical review. (LICOS Discussion Papers 18207) LICOS—Centre for Institutions and Economic Performance, K.U.Leuven.
van Biesebroeck, J. (2006). Exporting raises productivity in Sub-Saharan African manufacturing firms. Journal of International Economics, 67(2), 373–391.
van Biesebroeck, J. (2007). Robustness of productivity estimates. The Journal of Industrial Economics, 55(3), 529–569.
Blanes, J. V., & Busom, I. (2004). Who participates in R&D subsidy programs? The case of Spanish manufacturing firms. Research Policy, 33(10), 1459–1476.
Bustos, P. (2005). The impact of trade on technology and skill upgrading evidence from Argentina. Universitat Pompeu Fabra mimeo.
Bustos, P. (2009). Trade liberalization, exports and technology upgrading: Evidence on the impact of MERCOSUR on Argentinean firms. forthcoming The American Economic Review.
Campa, J. M. (1998). Hysteresis in trade: How big are the numbers? (Working Paper 9802) Programa de Investigaciones Económicas, Fundación Empresa Pública.
Carrasco, R. (2001). Binary choice with binary endogenous regressors in panel data: Estimating the effect of fertility on female labour participation. Journal of Business and Economics Statistics, 19(4), 385–394.
Cassiman, B., & Martínez-Ros, E. (2005). Product innovation and exports: Evidence from Spanish manufacturing. IESE mimeo.
Clerides, S., Lach, S., & Tybout, J. R. (1998). Is Learning by exporting important? Micro-dynamic evidence from Colombia, Mexico, and Morocco. Quarterly Journal of Economics, 113(3), 903–947.
Constantini, J. A., & Melitz, M. (2008). The dynamics of firm-level adjustment to trade liberalization. CEPR mimeo.
Das, S., Roberts, M. J., & Tybout, J. R. (2007). Market entry costs, producer heterogeneity, and export dynamics. Econometrica, 75(3), 837–873.
Delgado, M., Fariñas, J., & Ruano, S. (2002). Firm productivity and export markets: A non-parametric approach. Journal of International Economics, 57(2), 397–422.
de Loecker, J. (2007). Do exports generate higher productivity? Evidence from Slovenia. Journal of International Economics, 73(1), 69–98.
EU KLEMS. (2007). Productivity in the European Union; a comparative industry approach. EU KLEMS Database Project.
Fariñas, J. C., & Jaumandreu, J. (1999). Diez Años de la Encuesta sobre Estrategias Empresariales (ESEE). Economía Industrial (329), 29–42.
Fariñas, J. C., & Martín-Marcos, A. (2007). Exporting and economic performance firm-level evidence from manufacturing. The World Economy, 30(4), 618–646.
Fariñas, J. C., & Ruano, S. (2005). Firm productivity, heterogeneity, sunk costs and market selection. International Journal of Industrial Organization, 23(7–8), 505–534.
Girma, S., Görg, H., & Hanley, A. (2008). R&D and exporting: A comparison of British and Irish firms. Review of World Economics/Weltwirtschaftliches Archiv, 144(4), 750–773
Green, W. (2008). Econometric Analysis, 6th Edition. Upper Saddle River, NJ: Prentice Hall.
Greenaway, D., Guariglia, A., & Kneller, R. (2007). Do financial factors affect exporting decisions? Journal of International Economics, 73(2), 377–395.
Greenaway, D., Kneller, R. (2007). Firm heterogeneity, exporting and foreign direct investment. The Economic Journal, 117(517), F134–F161(28).
Griliches, Z. (1998). R&D and Productivity the Econometric Evidence. Chicago: University of Chicago Press.
Grossman, G. M., & Helpman, E. (1991). Innovation and growth in the global economy. Cambridge: MIT Press.
Guillén, M. F. (2005). The Rise of Spanish Multinationals: European Business in the Global Economy. Cambridge: Cambridge University Press.
Huergo, E., & Jaumandreu, J. (2004). Firms’ age, process innovation and productivity growth. International Journal of Industrial Organization, 22(4), 541–559.
Iacovone, L., & Javorcik, S. B. (2008). Getting ready: Preparation for exporting. University of Oxford mimeo.
Klette, T. J., & Griliches, Z. (1996). The inconsistency of common scale estimators when output prices are unobserved and endogenous. Journal of Applied Econometrics, 11(4), 343–61.
Konings, J., & Vandenbussche, H. (2008). Heterogeneous responses of firms to trade protection. Journal of International Economics, 76(2), 371–383.
Lileeva, A., & Trefler, D. (2007). Trade liberalization raises plant-level productivity...for some plants. (NBER Working Paper 13297)
Mairesse, J., & Jaumandreu, J. (2005). Panel data estimates of the production function and the revenue function: What difference does it make? Scandinavian Journal of Economics, 107(4), 651–672.
Mañez, J. A., Rochina-Barrachina, M. E., & Sanchis, J. A. (2008). Sunk cost hysteresis in Spanish manufacturing exports. Review of World Economics/Weltwirtschaftliches Archiv, 144(2), 272–294.
Markusen, J. R. (2002). Multinational Firms and the Theory of International Trade. Cambridge, MA: MIT Press.
Melitz, M. (2003). The impact of trade on intra-industry reallocations and aggregate industry productivity. Econometrica, 71(6), 1695–1725.
OECD (2004). Main science and technology indicators. Volume 2004/2.
Olley, S., & Pakes, A. (1996). The dynamics of productivity in the telecommunications equipments industry. Econometrica, 64(6), 1263–1298.
Peters, B. (2009). Persistence in innovation: Stylized facts and panel data evidence. The Journal of Technology Transfer, 34(2), 226–243.
Roberts, M., & Tybout, J. (1997). The decision to export in Colombia: An empirical model of entry with sunk costs. American Economic Review, 87(4), 545–564.
Trefler, A. (2004). The long and short of the Canada-US free trade agreement. American Economics Review, 94(4), 870–895.
Wagner, J. (2007). Exports and productivity: A survey of the evidence from firm-level data. The World Economy, 30(1), 60–82.
Yeaple, S. R. (2005). A simple model of firm heterogeneity, international trade, and wages. Journal of International Economics, 65(1), 1–20.
Acknowledgments
I am grateful to André Sapir for his support and constructive advice. I also thank Bee Yan Aw, Paola Conconi, Harald Fadinger, Victor Ginsburgh, Georg Kirchsteiger, Nicholas Sheard, Ziga Zarnic, two anonymous referees and seminar participants at ECARES-Université Libre de Bruxelles, the University College London ENTER Jamboree and the CEPR SCIFI-GLOW meeting in Madrid for useful discussions and suggestions. The financial support from the Fundación Ramón Areces and the Université Libre de Bruxelles for acquiring the data is gratefully acknowledged.
Author information
Authors and Affiliations
Corresponding author
Appendix: data appendix
Appendix: data appendix
1.1 Data cleaning
The data is cleaned from unlikely values, large spikes and missing values according to the following criteria. First, I dropped observations with negative real value added (305 of them). Second, I dropped the observations with unrealistically large spikes: a) Employment growth of more than 200%, when there was no merger in the previous or next year (38). b) Sales growth of more than 500% when there was no merger in the previous or next year (27). And c) Output growth of more than 500% when there was no merger in the pervious or next year (18). Third, I dropped firms that exhibit irregular exit patterns. That is firms for which one cannot be certain that they did not exit the sample before it was recorded in the data such as firms that i) do not answer or ii) do not collaborate in the year previous to exit the sample (5). Finally, I dropped observations with export intensity (ratio of exports to sales) larger than 1 (2). As a result of the cleaning process described above, less than 2% of observations are lost, so the sample remains highly representative of the population of Spanish manufacturing firms.
1.2 Definition of variables
All variables are measured annually at the firm level and are taken from the ESEE survey, unless otherwise stated. All monetary variables are in real terms and have been deflated using the appropriate deflators used below.
Age: age in a given year, computed as t minus the year of birth (as declared in the questionnaire) plus 1, since firms report the information at the end of the natural year.
Export dummy: equal to 1 if the firms reports positive export sales and 0 otherwise.
Export intensity: export sales over total sales.
Export sales: export sales as reported by the firm. Real exports are calculated deflating nominal values using the firm-specific output deflator described below.
Foreign ownership dummy: equal to 1 if the firm has a foreign participation share of equal to at least 20% of total capital, and 0 otherwise.
Industry dummies: 20 industry dummies. Each dummy takes the value 1 if the firm main activity is in that industry, and zero otherwise. See Table 2 for the industry breakdown.
R&D dummy: a dummy equal to 1 when the firm reports to conduct R&D activities and 0 when it reports not to do R&D.
R&D expenditure: total research and development (R&D) expenditure by the firm, including both intramural and extramural expenditures as defined by the Frascati manual, deflated using the consumer price index.
R&D intensity: R&D expenditure over total sales in percentage.
Product innovation dummy: a dummy equal to 1 when the firm reports to introduce a product innovation, and 0 otherwise. Product innovations are completely new products, or with such modifications that they are different from those produced earlier. For instance, such that it incorporates new materials, new intermediate products, new design, or new functions of the product.
Process innovation dummy: a dummy equal to 1 when the firm reports to introduce a process innovation, and 0 otherwise. For instance, changes in the process such as the introduction of new machinery, new methods of organizing the work, or both.
Public support for innovation dummy: a dummy equal to 1 when the firm reports to have received support for innovation from any public institution, which may include the Spanish central government, the regional governments, or other institutions like the European Union, and 0 otherwise.
Wage: average wage calculated as labor costs over the total number of employees.
Total factor productivity (TFP): is estimated using the Olley and Pakes (1996) employing the following variables. Ouput is measured by the sales minus production costs. Real output is constructed by deflating the nominal value using the firm-specific output deflator described below. Materials are measured by the cost of material inputs. Real material input is constructed deflating the nominal value using the firm-specific material input deflator described below. Capital is measured by total capital assets. Real capital is constructed by deflating the nominal value using the capital deflator described below. Investment is constructed using the perpetual inventory method and the capital variable. Real investment is constructed by deflating the nominal value using the investment deflator described below. And Labor is the number of workers multiplied by the number of hours per worker (normal hours of work plus overtime minus idle working hours).
1.3 Deflators
Output deflator: individual price indices for each firm are constructed using the information on output price changes in the firm’s main market drawn from the ESEE.
Material input deflator: individual price indices for each firm are constructed using information on the price changes and the costs of material inputs drawn from the ESEE; which include raw materials, energy and purchases of external services. The index is then constructed using the formula: \(P_{mat}=\frac{C_{serv}}{C_{mat}} P_{serv}+\frac{C_{raw+ener}}{C_{mat}}P_{raw+ener},\) where C serv stands for the cost of external services purchased by the firm, P serv is the price index of external services, Craw+ener is the cost of raw materials and energy, C mat represents the total cost of materials, and Praw+ener is the price index of raw materials and energy. Since the available data does not allow to distinguish between the relative weight of raw materials and energy, I took the geometric mean of both prices giving fixed weights to each component: Praw+ener = [P raw ]0.95[P ener ]0.05 following the procedure commonly used by the ESEE survey.
Capital deflator: average annual equipment goods component of the index of industry prices published by the Instituto Nacional de Estadística.
Investment deflator: average annual equipment goods component of the index of industry prices published by the Instituto Nacional de Estadística.
Producer price index (PPI): the gross output price indices at the industry-level are taken from the EU KLEMS database (March, 2007). Since the industrial classification used by the ESEE does not exactly match the one used by the EU KLEMS database, I aggregate from the original 20 sectors in the ESEE to 17. I aggregate sectors: [151]-Production, processing and preserving of meat, [152–158, 160]-Food products and tobacco and [151]-Beverages that are merged together into “Food, Beverages and Tobacco”. And sectors [361]-Furniture and [362–633; 371–372]-Other manufacturing into “Manufacturing nec and recycling”. The other sectors remain the same. The gross output price indices for the aggregated sectors are constructed by weighting the individual sector indices with their output shares over the aggregated sector output. To compute the weights I use information on gross output at current basic prices from EU KLEMS 2007. For the aggregated sectors, I then use information on gross output at current basic prices from EU KLEMS to compute the shares of the individual sectors over the aggregated sectoral output. Those shares are used to weight each of the individual output price index to obtain an output deflator at the aggregated sector level.
Consumer price index (CPI): average annual general consumer price index published by the Instituto Nacional de Estadística.
About this article
Cite this article
Caldera, A. Innovation and exporting: evidence from Spanish manufacturing firms. Rev World Econ 146, 657–689 (2010). https://doi.org/10.1007/s10290-010-0065-7
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10290-010-0065-7