Abstract
This paper uses a census of Russian manufacturing firms to study the relationship between exports and productivity at the firm level. The period studied, 1996–2002, implies that the results are affected by the Russian financial crisis of 1998. Exporters are shown to be more productive and larger than non-exporters, seemingly an effect of more productive firms self-selecting into the export market, rather than learning effects. But learning effects are significant among new entrants. Additionally, in examining the effect of the direction of exports on productivity, the finding is that the difference in the productivity level of firms exporting to the OECD and the CIS is insignificant.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
For a discussion on Russian industry restructuring and productivity see Ahrend (2004).
In 2003 the export share of petroleum, petroleum products and Gas (SITC, rev 3 code 33, 34 respectively) exceeded 50% (Ahrend 2004).
Roberts and Tybout (1997), for example, find empirical support for the existence of sunk entry costs.
For examples of representative-firm models see Helpman and Krugman (1985).
For a summary of empirical studies on export and productivity of firms/plants, see Wagner (2007).
Transfer pricing and price regulations make it difficult to assess the productivity in the electricity and fuel sectors. The metal sector is heavily dependent on natural resources and the chemical sector is linked to the oil industry.
For a more detailed discussion of the databases and their limitations see Bessonova et al. (2003).
Firms with less than 25 employees all years, with negative costs as percent of sales, with larger exports than total sales, with more workers than total number of employees, with costs more than 20,000% of sales have been excluded from the sample.
GNOZIS is a database, which includes statistical and balance sheet information on Russian firms. Coverage of firms in GNOZIS is similar to that in the Russian Enterprise Longitudinal Database.
The use of intermediates when conducting exports is rather common in Russia, as pointed out by seminar participants at CEFIR. Unfortunately, we cannot control for that, hence we prefer not to use export-shares, but rather a simple dummy variable indicating export status.
Calculations of the productivity measures and real value added are described in the appendix.
We have tried including other firm characteristics such as workers’ share of total employment and K/L-ratio, but the estimated coefficients were non-significant and the data quality is poor. Therefore, they have been omitted in our selected specification.
The random effect model was rejected by standard Hausman tests against the fixed effect model.
Since the standard transformation is likely to be biased, we apply the formula suggested by van Garderen and Shah (2002) to derive the % effect of the relevant dummy on the dependent variable and its standard errors: \( \hat{p}_{i} = 100{\left[ {\exp {\left( {\hat{c}_{i} - \frac{1} {2}\hat{v}{\left( {c_{i} } \right)}} \right)} - 1} \right]}{\text{ and }}\hat{v}{\left( {\hat{p}_{i} } \right)} = 100^{2} \exp {\left( {2\hat{c}_{i} } \right)}{\left[ {\exp {\left( { - \hat{v}{\left( {\hat{c}_{i} } \right)}} \right)} - \exp {\left( { - 2\hat{v}{\left( {\hat{c}_{i} } \right)}} \right)}} \right]} \) where \( \hat{p}_{i} \) is the transformed coefficient to calculate; \( \hat{c}_{i} \) is the estimated coefficient belonging to the relevant dummy variable; \( \hat{v}{\left( {\hat{c}_{i} } \right)} \) is the estimated variance of the same dummy variable.
The estimated export premium (TFP) in Sweden is less than 11.1% (Hansson and Lundin 2004).
The estimated export premium (Labour productivity) in Slovenia is slightly less than 30%.
Other variables such as average wage and K/L-ratios have been excluded since they have been found to be insignificant in the fixed effect estimations.
A description of the number of observations of each firm type in the periods used can be found in the appendix, Table A1.
The results are shown in the appendix, Table A5.
See for example Ghosal (2003).
Our analysis of the relationship between export composition and learning-by-exporting encountered several problems and the issue merits further research, but that goes beyond the scope of this paper. Note that firms classified in raw material sectors are excluded from our sample.
Labour productivity is higher in firms exporting mainly to the OECD than in firms exporting mainly to the CIS. The results are available from the corresponding author on request.
Differences in the use of barter transactions between exporters to the CIS and the OECD may affect our estimated TFP, leading to a downward bias of CIS exporters’ TFP, assuming that barter transactions are more frequent in intra-CIS trade. We thank Pertti Haaparanta for drawing our attention to this issue.
References
Ahrend R (2004) Accounting for Russia’s Post-crisis Growth. OECD Economics Department Working Papers No. 404, OECD Publishing, http://caliban.sourceoecd.org/vl=1596894/cl=24 /nw=1/rpsv/cgi-bin/wppdf?file=5lgsjhvj75wd.pdf
Alvarez R, Lopez R (2004) Exporting and performance: evidence from Chilean plants. Can J Econ 38(4):1384–1400
Bernard AB, Jensen J (1999) Exceptional exporter performance: cause, Effect or both? J Int Econ 47(1):1–25
Bernard AB, Jensen J (2004) Export and productivity in the USA. Oxford Rev Econ Policy 20(3):343–357
Bessonova E, Kozlov K, Yudaeva K (2003) Trade liberalization, Foreign Direct Investment, and Productivity of Russian Firms. CEFIR Working paper No. 39, New Economic School, Moscow, http://www.cefir.org/download.php?id=81
Blalock G, Gertler P (2004) Learning from exporting revisited in a less developed setting. J Dev Econ 75(2):397–416
Bleaney M, Filatotchev I, Wakelin K (2000) Learning by exporting: evidence from three transition economies. GLM Research Paper 2000/6, Centre for Research on Globalisation and Labour Markets, School of Economics, University of Nottingham, http://www.nottingham.ac.uk/ economics/leverhulme/research_papers/00_6.pdf
Castellani D (2002) Export behaviour and productivity growth: evidence from Italian manufacturing firms. Rev World Econ 138(4):605–628
Clerides S, Lach S, Tybout JR (1998) Is learning-by-exporting important? Micro-dynamic evidence from Colombia, Mexico, and Morocco. Q J Econ 113(3):903–947
Damijan JP, Polanec S, Prasniak J (2004) Self-selection, Export market heterogeneity and productivity improvements: firm level evidence from Slovenia. LICOS Discussion Paper 148/2004, LICOS Centre for Transition Economics, Katholieke Universiteit Leuven, Leuven http://www.econ.kuleuven.ac.be/licos/DP/DP2004/DP148.pdf
De Loecker J (2007) Do exports generate higher productivity? Evidence from Slovenia. J Int Econ 73(1):69–98
Djankov S, Hoekman B (1997) Productivity growth in Bulgarian enterprises. World Bank Policy Research Working paper 1707, The World Bank, Washington, DC, http://www.worldbank.org/research/trade/pdf/ wp1707.pdf
Ghosal V (2003) Impact of uncertainty and sunk cost on firm survival and industry dynamics. WZB Working paper SP II 2003-12, Social Science Research Center Berlin, http://skylla.wz-berlin.de/pdf/2003/ii03-12.pdf
Girma S, Greenaway D, Kneller R (2004) Does exporting increase productivity? A microeconometric analysis of matched firms. Rev Int Econ 12(5):855–866
Goskomstat (various issues) ‘Promyshlennost Rossii’, Moscow: Russian Statistical Agency
Hansson P, Lundin NN (2004) Exports as an indicator on or promoter of successful Swedish manufacturing firms in the 1990s. Rev World Econ 140(3):415–445
Helpman E, Krugman P (1985) Market structure and foreign trade: increasing returns, imperfect competition, ant the international economy. MIT press, Cambridge, MA
Helpman E, Melitz MJ, Yeaple SR (2004) Export versus FDI with heterogeneous firms. Am Econ Rev 94(1):300–316
Honoré BE, Kyriazidou E (2000) Panel data discrete choice models with lagged dependent variables. Econometrica 68(4):839–874
Melitz MJ (2003) The impact of trade on intra-industry reallocations and aggregate industry productivity. Econometrica 71(6):1695–1725
Roberts MJ, Tybout JR (1997) The decision to export in Colombia: an empirical model of entry with sunk costs. Am Econ Rev 87(4):545–564
Roodman D (2005) xtabond2: stata module to extend xtabond dynamic panel data estimator. Center for Global Development, Washington http://econpapers.repec.org/software/bocbocode /s435901.htm
Wagner J (2007) Exports and productivity: a survey of the evidence from firm level data. World Econ 30(1):60–82
van Garderen KJ, Shah C (2002) Exact interpretation of dummy variables in semilogarithmic equations. Econometr J 5(1):149–159
Acknowledgements
We would like to thank two anonymous reviewers for helpful comments and suggestions. Previous versions of this paper was presented at the NOITS 8th annual workshop in Helsinki, CEFIR and Lund University and the authors would like to thank the participants for helpful comments. We are indebted to Yves Bourdet, Joakim Gullstrand and Christian Jörgensen for valuable comments. Fredrik gratefully acknowledges financial support from The Crafoord foundation, Stiftelsen för främjande av ekonomisk forskning vid Lunds universitet and Sparbanksstiftelsen Färs & Frosta and the help of Guido Friebel.
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
1.1 Estimation of the TFP
First, value added is calculated as: sales - costs + wage bill. Value added is deflated by the output price index at the 5 digit industry level. Second, we estimate a Cobb-Douglas production function on 3-digit industry level (about 28 industries) or a translog production function on 2-digit industry level (6-industries) using one-way fixed effects. TFP is then calculated as the difference between the observed and predicted value for each firm and time period.
1.2 Results using the translog production function to estimate the productivity
Rights and permissions
About this article
Cite this article
Wilhelmsson, F., Kozlov, K. Exports and productivity of Russian firms: in search of causality. Econ Change 40, 361–385 (2007). https://doi.org/10.1007/s10644-008-9038-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10644-008-9038-4