Skip to main content
SpringerLink
Log in

FDI spillover effects in incomplete datasets

  • Article
  • Published:
Journal of International Business Studies Aims and scope Submit manuscript

Abstract

Scholars studying foreign direct investment (FDI) spillovers usually examine whether productivity gains in domestic firms can be attributed to the presence of foreign firms in their industry. However, empirical estimation is often based on datasets that omit certain kinds of firms in the economy. We argue that identifying FDI spillover effects in such incomplete datasets is problematic, owing to measurement error and selection problems. Using Monte Carlo simulations, we show that spillover effect estimates from incomplete datasets are potentially biased. We discuss the theoretical implications of this, and demonstrate a weighted instrumental variable approach that could yield better spillover effect estimates in incomplete datasets.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Figure 1

Similar content being viewed by others

Notes

  1. Identification, in general, pertains to whether there is enough information available to allow a model to converge to one unique set of values for the parameters. If, say, two models or sets of parameters fit the data equally well, we usually cannot tell which one of the two we are estimating. In this sense, both models are not identified. But fundamentally, since the parameters of interest in the model usually represent causal relationships, identifying them means being able to trace out the causal relationships from the available data (e.g., Angrist & Krueger, 2001; Angrist & Pischke, 2009; Manski, 1995). So identifying the spillover effect basically means ensuring that the theoretical model we specify and estimate reflects the true spillover effect (cf. Koopmans, 1949; Manski, 1995). We explain this further later in the paper.

  2. Identification concerns in the literature this far have been around simultaneity of domestic firm productivity and foreign presence (Keller, 2009), selection effects from not observing firms after they exit (Haskel et al., 2007), and input endogeneity in the production function that could bias the estimates of input elasticities (Castellani & Zanfei, 2006; Levinsohn & Petrin, 2003; Marschak & Andrews, 1944). Identification challenges in incomplete datasets due to measurement error and selection problems have not been formally examined. Haskel et al. (2007) and Keller and Yeaple (2009) do discuss measurement error, but not as an identification problem. Moreover, they take the view that measurement error always biases the spillover effect estimates downward, and that instrumental variable regression overcomes measurement error issues. We will show later in the paper that neither is necessarily true. Also, while selection issues discussed in the literature pertain to survival biases arising from not being able to observe firms after they exit, our selection issue pertains to completely censored data on small firms, which, we will argue, poses very difficult identification challenges.

  3. The Prowess database does contain other firms as well, but as Marin and Sasidharan (2010: 1231) themselves say, these tend to be large public sector enterprises.

  4. In fact, Koopmans (1949), who introduced the term “identification” into the econometric literature, puts it this way: “Statistical inference from observations to economic behaviour parameters can be made in two steps: inference from the observations to the parameters of the assumed joint distribution of the observations, and inference from that distribution to the parameters of the structural equations describing economic behaviour. The latter problem of inference (is) described by the term identification problem” (125). So, for identification, the assumed joint distribution of observations (i.e., the theoretical model) should adequately reflect the set of structural equations that describe the true spillover process.

  5. If we express the relationship between mismeasured foreign presence (fp j ) and true foreign presence (FP j ) as a linear function such as: γ0+γ1FP j +ξ, where ξ is an error term independent of FP j , and has 0 mean and constant variance, then what the estimated spillover effect θ1 is really identifying is a term λβ1, where λ=γ1Var(FP j )/[γ12Var(FP j )+Var(U j )]. In this case, it is possible for the estimated effect θ1 to overestimate the true effect β1 as long as γ1≠1 (Carroll et al., 2006: 47).

  6. Secondary datasets, as we have seen, tend to be biased towards publicly listed firms, and are thus likely to miss privately held foreign and domestic firms. Missing foreign firms can lead to measurement error, as we have seen in the previous section, but it should not raise selection problems. This is because the spillover specification is usually estimated only on the domestic firms in the dataset. Missing (privately held) domestic firms, on the other hand, can lead to selection problems if the missing domestic firms also tend to be the smaller ones in the economy.

  7. We thank an anonymous reviewer for making this point.

  8. The Chinese dataset, as we have indicated before, does have a size bias. However, this is not a problem, given that the objective of our simulation is not to assess FDI spillovers in China. Instead, we want to examine what happens when, assuming the Chinese dataset is the full population, spillover effects are estimated from subsamples of the dataset that are systematically incomplete. If our objective was to estimate spillover effects for China, then, indeed, the Chinese dataset might be inappropriate, given its size bias.

  9. Ideally, we should examine the effect of missing privately held firms and small firms, because these mirror the specific ways in which secondary datasets and manufacturing census surveys that have been used in the past are biased. But from the Chinese NBS data that we had access to, although we had data on firm size, we did not know whether each firm was publicly listed or not. Hence, as the next best alternative, we examined the effects on estimated spillover effects of missing foreign firms. The logic is that if datasets are biased towards public firms, then it is quite likely that foreign wholly owned ventures are missing. So the case of missing foreign firms we examine is a subset of the bigger problem of datasets being biased towards publicly listed firms.

  10. We initially assume a true spillover effect of +2, but in later robustness tests replace this with several other positive and negative values drawn from normal as well as alternative underlying distributions.

  11. If we wanted to allow for selection effects as well, we would use different distributions for different size categories of firms. We would assign spillover effects as draws from a distribution with a higher mean, say 3, for large firms that are better able to absorb spillovers, and from a normal distribution with smaller mean, say 0, for small firms that lack absorptive capacity. In this particular simulation, however, we wanted to keep effects of selection effects constant, and hence drew spillover effect values for all firms from the same distribution.

  12. We calculated the dependent variable, productivity of domestic firms, as follows. We first estimated a Cobb–Douglas production function using the Levinsohn and Petrin (2003) routine for each of the industries in our sample, and obtained industry-specific coefficients of capital and labor. Then, for each domestic firm, we calculated the residual, that is, the difference between the firm’s actual output and its expected output, given its usage of capital and labor. This residual, as in most prior FDI spillover studies, is our measure of total factor productivity (Arnold, 2005).

  13. We thank an anonymous reviewer for suggesting that we run these simulations.

  14. These results pertain to the asymptotic, large sample properties of this estimator. In finite samples, there might still be some bias even, though the estimator is consistent in infinite samples. This is called the “finite-sample bias of IV” (Cameron & Trivedi, 2009: 176).

  15. We thank an anonymous reviewer for pointing out that it is incorrect for us to claim that a specific weighting scheme that worked in our dataset will work equally well for scholars using other datasets. Since the extent of size bias varies from dataset to dataset, the only way to know which weight to use, we think, is to compare them in model selection tests, as we describe in our empirical demonstration.

References

  • Acemoglu, D., & Dell, M. 2010. Productivity differences between and within countries. American Economic Journal: Macroeconomics, 2 (1): 169–188.

    Google Scholar 

  • Aiken, L. S., & West, S. G. 1991. Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage Publications.

    Google Scholar 

  • Aitken, B. J., & Harrison, A. E. 1999. Do domestic firms benefit from direct foreign investment? Evidence from Venezuela. The American Economic Review, 89 (3): 605–618.

    Article  Google Scholar 

  • Altomonte, C., & Pennings, E. 2009. Domestic plant productivity and incremental spillovers from foreign direct investment. Journal of International Business Studies, 40 (7): 1131–1148.

    Article  Google Scholar 

  • Angrist, J., & Krueger, A. B. 2001. Instrumental variables and the search for identification: From supply and demand to natural experiments. Journal of Economic Perspectives, 15 (4): 69–85.

    Article  Google Scholar 

  • Angrist, J., & Pischke, J.-S. 2009. Mostly harmless econometrics. Princeton, NJ: Princeton University Press.

    Google Scholar 

  • Arnold, J. M. 2005. Productivity estimation at the plant level: A practical guide. Unpublished manuscript. Available at http://www.jensarnold.de.

  • Bae, J., & Coo, J. 2008. Information loss, knowledge transfer cost and the value of social relations. Strategic Organization, 6 (3): 227–258.

    Article  Google Scholar 

  • Baily, M. N., & Solow, R. M. 2001. International productivity comparisons built from the firm level. Journal of Economic Perspectives, 15 (3): 151–172.

    Article  Google Scholar 

  • Banga, R. 2006. The export-diversifying impact of Japanese and US foreign direct investments in the Indian manufacturing sector. Journal of International Business Studies, 37 (4): 558–568.

    Article  Google Scholar 

  • Bartelsman, E. J., Haltiwanger, J. C., & Scarpetta, S. 2013. Cross-country differences in productivity: The role of allocation and selection. American Economic Review, 103 (1): 305–334.

    Article  Google Scholar 

  • Becker, R.A., & Gray, W.B. 2009. NBER-CES Manufacturing Industry Database. Available from http://www.nber.org/data/nbprod2005.html, accessed 9 February 2011.

  • Bernard, A. B., & Jones, C. I. 1996. Comparing apples to oranges: Productivity convergence and measurement across industries and countries. American Economic Review, 86 (5): 1216–1238.

    Google Scholar 

  • Blalock, G., & Simon, D.H. 2009. Do all firms benefit equally from downstream FDI? The moderating effect of local suppliers' capabilities on productivity gains. Journal of International Business Studies, 40 (7): 1095–1112.

    Article  Google Scholar 

  • Bound, J., Brown, C., & Mathiowetz, N. 2001. Measurement error in survey data. In J. J. Heckman, & E. Leamer (Eds), Handbook of econometrics, 3705–3843. Amsterdam, The Netherlands: Elsevier.

    Google Scholar 

  • Buckley, P. J., & Casson, M. 1976. The future of the multinational enterprise. London: Palgrave Macmillan.

    Book  Google Scholar 

  • Cameron, A. C., & Trivedi, P. K. 2009. Microeconometrics using Stata. College Station, TX: Stata Press.

    Google Scholar 

  • Cantner, U., & Pyka, A. 1998. Absorbing technological spillovers: Simulations in an evolutionary framework. Industrial and Corporate Change, 7 (2): 369–397.

    Article  Google Scholar 

  • Carroll, R. J., Ruppert, D., Stefanski, L. A., & Crainiceanu, C. M. 2006. Measurement error in nonlinear models: A modern perspective. London, UK: Chapman & Hall.

    Book  Google Scholar 

  • Castellani, D., & Zanfei, A. 2006. Multinational firms, innovation and productivity. Cheltenham: Edward Elgar.

    Book  Google Scholar 

  • Castellani, D., & Zanfei, A. 2007. Multinational companies and productivity spillovers: Is there a specification error? Applied Economic Letters, 14 (14): 1047–1051.

    Article  Google Scholar 

  • Caves, R. E. 1996. Multinational enterprise and economic analysis. Cambridge, UK: Cambridge University Press.

    Google Scholar 

  • Centola, D., & Macy, M. 2007. Complex contagions and the weakness of long ties. American Journal of Sociology, 113 (3): 702–734.

    Article  Google Scholar 

  • Costello, D. M. 1993. A cross-country, cross-industry comparison of productivity growth. Journal of Political Economy, 101 (2): 207–222.

    Article  Google Scholar 

  • Chang, S.-J., & Xu, D. 2008. Spillovers and competition among foreign and local firms in China. Strategic Management Journal, 29 (5): 495–518.

    Article  Google Scholar 

  • Chen, X., Hong, H., & Nekipelov, D. 2011. Nonlinear models of measurement errors. Journal of Economic Literature, 49 (4): 901–937.

    Article  Google Scholar 

  • Chung, W., Mitchell, W., & Yeung, B. 2003. Foreign direct investment and host country productivity: The American automotive component industry in the 1980s. Journal of International Business Studies, 34 (2): 199–218.

    Article  Google Scholar 

  • Dechow, P. M. 1994. Accounting earnings and cash flows as measures of firm performance: The role of accounting accruals. Journal of Accounting and Economics, 18 (1): 3–42.

    Article  Google Scholar 

  • Driffield, N., & Jindra, B. 2012. Challenging the production function approach to assess the developmental effects of FDI. European Journal of Development Research, 24 (1): 32–37.

    Article  Google Scholar 

  • Driffield, N., & Love, J. H. 2007. Linking FDI motivation and host economy productivity effects: Conceptual and empirical analysis. Journal of International Business Studies, 38 (3): 460–473.

    Article  Google Scholar 

  • Driffield, N., Love, J. H., & Menghinello, S. 2010. The multinational enterprise as a source of international knowledge flows: Direct evidence from Italy. Journal of International Business Studies, 41 (2): 350–359.

    Article  Google Scholar 

  • Dumouchel, W. H., & Duncan, G. J. 1983. Using sample survey weights in multiple regression analyses of stratified samples. Journal of the American Statistical Association, 78 (383): 535–543.

    Article  Google Scholar 

  • Eapen, A. 2012. Social structure and technology spillovers from foreign to domestic firms. Journal of International Business Studies, 43 (3): 244–263.

    Article  Google Scholar 

  • Fuller, W. A. 1987. Measurement error models. New York: John Wiley.

    Book  Google Scholar 

  • Gorg, H., & Strobl, E. 2001. Multinational companies and productivity spillovers: A meta-analysis. The Economic Journal, 111 (475): F723–F739.

    Article  Google Scholar 

  • Griffith, R., Redding, S., & Simpson, H. 2003. Productivity convergence and foreign ownership at the establishment level. CEPR Discussion Paper No. 3765, London, UK, Centre for Economic Policy Research.

  • Growiec, J., Pammolli, F., Riccaboni, M., & Stanley, H. E. 2008. On the size distribution of business firms. Economics Letters, 98 (2): 207–212.

    Article  Google Scholar 

  • Hall, R. E., & Jones, C. I. 1999. Why do some countries produce so much more output per worker than others? The Quarterly Journal of Economics, 114 (1): 83–116.

    Article  Google Scholar 

  • Harrigan, J. 1999. Estimation of cross-country differences in industry production functions. Journal of International Economics, 47 (2): 267–293.

    Article  Google Scholar 

  • Harrison, J. R., Lin, Z., Carroll, G. R., & Carley, K. M. 2007. Simulation modeling in organizational and management research. Academy of Management Review, 32 (4): 1229–1245.

    Article  Google Scholar 

  • Haskel, J. E., Pereira, S., & Slaughter, M. J. 2007. Does inward foreign direct investment boost the productivity of domestic firms? Review of Economics and Statistics, 89 (3): 482–496.

    Article  Google Scholar 

  • Hausman, J. 2001. Mismeasured variables in econometric analysis: Problems from the right and problems from the left. Journal of Economic Perspectives, 15 (4): 57–67.

    Article  Google Scholar 

  • Hennart, J.-F. 1982. A theory of multinational enterprise. Ann Arbor, MI: University of Michigan Press.

    Google Scholar 

  • Horowitz, J. L., & Manski, C. 1998. Censoring of outcomes and regressors due to survey nonresponse: Identification and estimation using weights and imputations. Journal of Econometrics, 84 (1): 37–58.

    Article  Google Scholar 

  • Hsieh, C.-T., & Klenow, P. J. 2009. Misallocation and manufacturing TFP in China and India. The Quarterly Journal of Economics, 124 (4): 1403–1448.

    Article  Google Scholar 

  • Hymer, S. H. 1976. The international operations of national firms: A study of direct foreign investment. Boston, MA: MIT Press.

    Google Scholar 

  • Javorcik, B. S. 2004. Does foreign direct investment increase the productivity of domestic firms? In search of spillovers through backward linkages. The American Economic Review, 94 (3): 605–627.

    Article  Google Scholar 

  • Keller, W. 2009. International trade, foreign direct investment, and technology spillovers. National Bureau of Economic Research Working Paper Series no. 15442.

  • Keller, W., & Yeaple, S. 2009. Multinational enterprises, international trade, and productivity growth: Firm-level evidence from the United States. Review of Economics and Statistics, 91 (4): 821–831.

    Article  Google Scholar 

  • King, R. G., & Levine, R. 1993. Finance and growth: Schumpeter might be right. The Quarterly Journal of Economics, 108 (3): 717–737.

    Article  Google Scholar 

  • Kipnis, V., Carroll, R. J., Freedman, L. S., & Li, L. 1999. Implications of a new dietary measurement error model for estimation of relative risk: Application to four calibration studies. American Journal of Epidemiology, 150 (6): 642–651.

    Article  Google Scholar 

  • Kish, L. 1990. Weighting: Why, when, and how? Proceedings of the Survey Research Methods Section, American Statistical Association, 121–130.

  • Koopmans, T. C. 1949. Identification problems in economic model construction. Econometrica, 17 (2): 125–144.

    Article  Google Scholar 

  • Krasker, W. S., & Welsch, R. E. 1985. Resistant estimation for simultaneous-equations models using weighted instrumental variables. Econometrica, 53 (6): 1475–1488.

    Article  Google Scholar 

  • Levinsohn, J., & Petrin, A. 2003. Estimating production functions using inputs to control for unobservables. Review of Economic Studies, 70 (2): 317–342.

    Article  Google Scholar 

  • Manski, C. F. 1995. Identification problems in the social sciences. Cambridge, MA: Harvard University Press.

    Google Scholar 

  • Marin, A., & Sasidharan, S. 2010. Heterogeneous MNC subsidiaries and technological spillovers: Explaining positive and negative effects in India. Research Policy, 39 (9): 1227–1241.

    Article  Google Scholar 

  • Marschak, J., & Andrews, W. 1944. Random simultaneous equations and the theory of production. Econometrica, 12 (3/4): 143–205.

    Article  Google Scholar 

  • Martin, X. 2013. Solving theoretical and empirical conundrums in international strategy research: Linking foreign entry mode choices and performance. Journal of International Business Studies, 44 (1): 28–41.

    Article  Google Scholar 

  • Meagher, K., & Rogers, M. 2004. Network density and R&D spillovers. Journal of Economic Behavior & Organization, 53 (2): 237–260.

    Article  Google Scholar 

  • Meyer, K. E., & Sinani, E. 2009. When and where does foreign direct investment generate positive spillovers? A meta-analysis. Journal of International Business Studies, 40 (7): 1075–1094.

    Article  Google Scholar 

  • Mooney, C. Z. 1997. Monte Carlo simulation. London, UK: Sage Publications.

    Book  Google Scholar 

  • Murray, M. P. 2006. Avoiding invalid instruments and coping with weak instruments. Journal of Economic Perspectives, 20 (4): 111–132.

    Article  Google Scholar 

  • OECD. 2010. STAN R&D: Research and Development Expenditure in Industry. STAN: OECD Structural Analysis Statistics (database). Available at http://dx.doi.org/10.1787/data-00032-en, accessed 9 February 2011.

  • Oxley, J., Rivkin, J. W., & Ryall, M. D. 2010. The strategy research initiative: Recognizing and encouraging high-quality research in strategy. Strategic Organization, 8 (4): 377–386.

    Article  Google Scholar 

  • Pfefferman, D. 1993. The role of sampling weights when modeling survey data. International Statistical Review, 61 (2): 317–337.

    Article  Google Scholar 

  • Pfefferman, D. 1996. The use of sampling weights for survey data analysis. Statistical Methods in Medical Research, 5 (3): 239–261.

    Article  Google Scholar 

  • Porter, R. D. 1973. On the use of survey sample weights in the linear model. Annals of Economic and Social Measurement, 2 (2): 140–157.

    Google Scholar 

  • Rajan, R. G., & Zingales, L. 1998. Financial dependence and growth. American Economic Review, 88 (3): 559–586.

    Google Scholar 

  • Reeb, D., Sakakibara, M., & Mahmood, I. P. 2012. From the editors: Endogeneity in international business research. Journal of International Business Studies, 43 (3): 211–218.

    Article  Google Scholar 

  • Reichstein, T., & Jensen, M. B. 2005. Firm size and firm growth rate distributions? The case of Denmark. Industrial and Corporate Change, 14 (6): 1145–1166.

    Article  Google Scholar 

  • Shaver, M. Y. 1998. Accounting for endogeneity when assessing strategy performance: Does entry mode choice affect FDI survival? Management Science, 44 (4): 571–585.

    Article  Google Scholar 

  • Simon, H. A., & Bonini, C. P. 1958. The size distribution of business firms. American Economic Review, 48 (4): 607–617.

    Google Scholar 

  • Smeets, R., & de Vaal, A. 2011. Knowledge diffusion from FDI and intellectual property rights. CPB Discussion paper no. 168, CPB Netherlands Bureau for Economic Research, The Hague.

  • Visek, J. A. 2009. Consistency of the instrumental weighted variables. Annals of the Institute of Statistical Mathematics, 61 (3): 543–578.

    Article  Google Scholar 

  • Vuong, Q. H. 1989. Likelihood ratio tests for model selection and non-nested hypotheses. Econometrica, 57 (2): 307–333.

    Article  Google Scholar 

  • Zamborsky, P. 2012. Intangibles of host country effects of foreign direct investment. Journal of International Business and Economics, 12 (2): 123–137.

    Google Scholar 

  • Zhang, Y., Li, H., Li, Y., & Zhou, L.-A. 2010. FDI spillovers in an emerging market: The role of foreign firms’ country origin diversity and domestic firms’ absorptive capacity. Strategic Management Journal, 31 (9): 969–989.

    Google Scholar 

Download references

Acknowledgements

I thank the editor and three anonymous reviewers, Anthea Zhang, Anthony Obeyesekere, Dave McKendrick, Grzeg Trojanowski, Peter Zamborsky, Rajiv Krishnan, Rekha Krishnan, Roger Smeets, Shivani Sourindre, and seminar participants at the universities of Melbourne, Sydney, and Auckland for their comments. I particularly thank Gaurab Aryal for helpful discussions on identification and model selection, Sea-Jin Chang for access to Chinese National Bureau of Statistics data, and Joachim Mai from Intersect for facilitating access to their supercomputer to run the simulations.

Author information

Authors and Affiliations

  1. Research School of Management, Australian National University, Canberra, Australia

    Alex Eapen

Authors
  1. Alex Eapen
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Alex Eapen.

Additional information

Accepted by Mariko Sakakibara, Area Editor, 29 May 2013. This paper has been with the author for two revisions.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Eapen, A. FDI spillover effects in incomplete datasets. J Int Bus Stud 44, 719–744 (2013). https://doi.org/10.1057/jibs.2013.32

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1057/jibs.2013.32

Keywords

Search

Navigation