Skip to main content


Log in

Should I stay or should I go? Foreign direct investment, employment protection and domestic anchorage

  • Original Paper
  • Published:
Review of World Economics Aims and scope Submit manuscript


This paper examines theoretically and empirically how employment protection legislation affects location decisions of multinationals. We depart from the “conventional wisdom” by examining not only the effect of protection on inward foreign direct investment (FDI), but also a country’s ability to “anchor” potential outward investment. Based on our simple theoretical framework, we estimate an empirical model, using data on bilateral FDI and employment protection indices for OECD countries, and controlling for other labour market institutions and investment costs. We find that, while an “unfavourable” employment protection differential between a domestic and a foreign location is inimical to FDI, a high domestic level of employment protection tends to discourage outward FDI. The results are in line with our conjecture that strict employment protection in the firm’s home country makes firms reluctant to relocate abroad and keeps them “anchored” at home.

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

Access this article

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

Instant access to the full article PDF.

Similar content being viewed by others


  1. This quotes a chairman of a regional development agency in the UK as saying that “The big market now is in the retention of the investment business that is here”. The Observer, “Another week, another firm quits the UK”, by Oliver Morgan, 1 June 2003.

  2. Hiring and firing restrictions are typically not found to have a decisive role on overall rates of unemployment (e.g. Nickell 1998; Nickell et al. 2001), but are shown to reduce job reallocation rates and employment variation over the business cycle—e.g. Bentolila and Bertola (1990) and Garibaldi et al. (1997).

  3. The monopoly assumption is similar to the one in the model by Haaland et al. (2003). We make this assumption for simplicity as the main purpose of the model is to motivate our empirical analysis. For a theoretical analysis of the ‘attraction versus retention’ effects of employment protection in strategic settings, see Dewit et al. (2003).

  4. These assumptions allow us to focus on the effect of employment protection on location choice, while abstracting from other location determinants, such as market access and other aspects of labour market institutions whose importance for firms’ location decisions is well understood. See, for instance, Markusen (2002) and Leahy and Montagna (2000a).

  5. As shown in the Appendix, the qualitative conclusions from our analysis remain unaltered. The only difference is that, with the possibility of a negative demand shock in period two, the attractiveness of home as the firm’s initial location is further reduced.

  6. Clearly, the two countries could be taken to represent two generic locations considered by a firm with headquarters in a third country. In this instance, Φf > Φh simply captures that the FDI costs for country f are higher than those for country h.

  7. Given the cost function, a firm will not operate from multiple production locations at the same time. With alternative cost functions, partial relocation may occur, but the main message—employment protection makes relocation less likely—will be preserved.

  8. For the same reasons, a negative demand shock in period two will narrow the difference between operating profits attainable in ‘foreign’ and those attainable in ‘home’ and hence makes the ‘foreign’ location even less attractive in period two than in period one. Therefore, relocation will not occur when demand in period two falls (see Appendix).

  9. In fact, changes in parameters other than those related to period two demand may cause relocation to ‘foreign’ in period two. For instance, one could consider uncertainty on the cost side. If the marginal cost of production in ‘foreign’ fell in period two with a given probability, then—for a large enough cost reduction—relocation from ‘home’ to ‘foreign’ would be a possibility.

  10. Of course, if the firm expected that the employment protection level in h were to fall in period two, then this would increase the attractiveness of h in period one as the firm’s initial location.

  11. An important shortcoming of the data for the dependent variable is that it does not allow us to distinguish new locations and relocations. Unfortunately we do not have data available that would enable such a distinction.

  12. Critics may argue that such an index is likely to be subjective. However, the perceptions of the managers of the firm as to the ‘desirability’ of a location are likely to play a crucial role in determining their decision.

  13. See OECD (1991) for a detailed description.

  14. See Elmeskov et al. (1998) for a discussion of the wage coordination and unionisation variables.

  15. All nominal variables are converted into real 1995 US dollars.

  16. While the investment cost variable is defined as an index, the interaction terms are based on a dummy equal to 1 if investment costs are higher than the median. In preliminary regressions we also interacted the EP variables with the actual level of investment costs. However, this produced unsatisfactory results and is not pursued here. This suggests that there are important non-linearities in the relationship between EP and investment costs which are better captured by the dummy variable.

  17. The GMM estimator is preferable to a standard IV estimator in the presence of heteroscedasticity. Our estimations suggest that first and second lags are not valid instruments and hence use third and fourth lags.

  18. Of course, the coefficients obtained from these estimations need to be treated with caution as they may be biased due to excluding country-partner unobservable. However, this bias works in the same direction for all estimations and still allows us to check whether there are differences in using the GCR or OECD indices.

  19. The magnitude of the coefficients is somewhat different. This is to be expected as the absolute values of the indices are not comparable. What we are interested in is the variation in the indices, however.

  20. See Leahy and Montagna (2000b) for a theoretical analysis of the strategic use of unionisation laws.

  21. If the negative shock is small enough, Eq. (A.1b) and (A.2) show that \( q_{1}^{h} < q_{2}^{h} \) and hence I θ  = 0, without making any qualitative changes to our analysis.

  22. If η is small enough so that \( q_{1}^{h} < q_{2}^{h} , \) then I θ  = 0 and \( \bar{\Upphi }^{h} = (\rho + \theta )\Upphi^{f} - \rho \left[ {\frac{{(a - c^{f} )^{2} - (a - c^{h} )^{2} }}{4b}} \right] - \frac{{(a - c^{f} )^{2} - [(a - c^{h} ) - (1 - \rho - \theta )\lambda^{h} ]^{2} }}{4b} \) with, once again, \( d\bar{\Upphi }^{h} /d\lambda^{h} < 0. \)


  • Baum, C. F., Schaffer, M. E., & Stillman, S. (2003). Instrumental variables and GMM: Estimation and testing. Working Paper No. 545, Boston College.

  • Bentolilla, S., & Bertola, G. (1990). Firing costs and labour demand: How bad is eurosclerosis? Review of Economic Studies, 57, 381–402.

    Article  Google Scholar 

  • Blonigen, B. A. (1997). Firm-specific assets and the link between exchange rates and foreign direct investment. American Economic Review, 87, 447–465.

    Google Scholar 

  • Culem, C. G. (1988). The locational determinants of direct investments among industrialised countries. European Economic Review, 32, 885–904.

    Article  Google Scholar 

  • Dewit, G., Leahy, D., & Montagna, C. (2003). Employment protection and globalisation in dynamic oligopoly. CEPR Discussion Paper 3871.

  • Elmeskov, J., Martin, J., & Scarpetta, S. (1998). Key lessons for labour market reforms: Evidence from OECD countries’ experiences. Swedish Economic Policy Review, 5, 205–252.

    Google Scholar 

  • Garibaldi, P., Konings, J., & Pissaridis, C. (1997). Gross job reallocation and labour market policy. In D. Snower & G. de la Dehesa (Eds.), Unemployment policy: Government options for the labour market. Cambridge: Cambridge University Press.

    Google Scholar 

  • Görg, H. (2005). Fancy a stay at the Hotel California? The role of easy entry and exit for FDI. Kyklos, 58, 519–536.

    Article  Google Scholar 

  • Haaland, J. I., & Wooton, I. (2002). Multinational investment, industry risk and policy competition. CEPR Discussion Paper 3152.

  • Haaland, J. I., Wooton, I., & Faggio, G. (2003). Multinational firms: Easy come, easy go? FinanzArchiv, 59, 3–26.

    Article  Google Scholar 

  • Javorcik, B. S., & Spatareanu, M. (2005). Do foreign investors care about labor market regulations? Review of World Economics/Weltwirtschaftliches Archiv 141, 375–403.

    Article  Google Scholar 

  • King, M. A., & Fullerton, D. (1984). The taxation of income from capital. Chicago: University of Chicago Press.

    Google Scholar 

  • Leahy, D., & Montagna, C. (2000a). Unionisation, rent extraction and FDI: Challenging conventional wisdom? Economic Journal, 100, 80–92.

    Article  Google Scholar 

  • Leahy, D., & Montagna, C. (2000b). Temporary social dumping, union legalisation and FDI: A note on the strategic use of standards. Journal of International Trade and Economic Development, 9, 243–259.

    Article  Google Scholar 

  • Markusen, J. R. (2002). Multinational firms and the theory of international trade. Cambridge, MA: MIT Press.

    Google Scholar 

  • Nickell, S. (1998). Unemployment, questions and some answers. Economic Journal, 108, 802–816.

    Article  Google Scholar 

  • Nickell, S., Nunziata, L., & Quintini, G. (2001). The beveridge curve, unemployment and wages in the OECD from the 1960s to the 1990s. CEP Discussion Paper 502, London School of Economics.

  • Nicoletti, G., Scarpetta, S., & Boylaud O. (1999). Summary indicators of product market regulation with an extension to employment protection legislation. OECD Economics Department Working Paper No. 226.

  • Nicoletti, G., Golub, S., Hajkova, D., Mirza D., & Yoo K.-Y. (2003). Policies and international integration: Influences on trade and foreign direct investment. OECD Economics Department Working Paper No. 359.

  • OECD (1991). Taxing profits in a global economy. Paris: OECD.

    Google Scholar 

  • Wheeler, D., & Mody, A. (1992). International investment location decisions: The case of U.S. firms. Journal of International Economics, 33, 57–76.

    Article  Google Scholar 

Download references


We are grateful to Dana Hajkova and Daniel Mirza for help with the OECD data and to David Greenaway, Dermot Leahy, Hassan Molana and an anonymous referee for helpful comments. We also gratefully acknowledge financial support from The Leverhulme Trust under Programme Grant F114/BF and the European Commission under Grant No. SERD-2002-00077.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Holger Görg.



In this appendix, we extend the basic model developed in the text by incorporating the possibility of a negative demand shock. Now, there are three possible states for period two demand: with probability ρ, demand in t = 2 is the same as in t = 1; with probability θ period-two demand will fall, i.e. p 2 = a − bq 2 − η, with η (a constant parameter) denoting the fall in demand; the probability that demand in the period two will boom is now 1 − ρ − θ.

  1. (i)

    Period two: the relocation decision

Equations (1)–(4b) remain valid, but we now have an additional expression for \( q_{2}^{f} \) and for \( q_{2}^{h} \) in the third possible state of period two demand, with:

$$ q_{2}^{f} = (a - c^{f} - \eta )/2b\quad {\text{if}}\quad p_{2} = a - bq_{2} - \eta $$


$$ q_{2}^{h} = (a - c^{h} - \eta + I_{\theta } \lambda^{h} )/2b\quad {\text{if}}\quad p_{2} = a - bq_{2} - \eta $$

I θ is an indicator variable with I θ  = 1 if \( q_{1}^{h} > q_{2}^{h} \left| {_{{p_{2} = a - bq_{2} - \eta }} } \right. \) and I θ  = 0 otherwise. When demand in t = 2 is the same as in t = 1, the firm will choose the same location in t = 2 as in t = 1; hence, no relocation will occur in that case. So, if period two demand is lower than demand in period one, the monopolist will a fortiori choose the same location as in period one (since maximised profits are convex in output). Therefore, relocation will only occur if the demand shock in period two is a positive one and hence Eq. (5) remains valid and is not altered by the possibility of a negative period two demand shock. Of course, the probability that relocation will happen is now 1 − ρ − θ instead of 1 − ρ.

  1. (ii)

    Period one: the initial location decision

We first calculate (h 1). When choosing \( q_{1}^{h} , \) the firm maximises \( E\pi (h_{1} ) = \pi_{1}^{h} + E\pi_{2}^{{h_{1} }} , \) but now \( E\pi_{2}^{{h_{1} }} = \rho \pi_{2}^{{h_{1} h_{2} }} \left| {_{{p_{2} = a - bq_{2} }} } \right. + \theta \pi_{2}^{{h_{1} h_{2} }} \left| {_{{p_{2} = a - bq_{2} - \eta }} } \right. + (1 - \rho - \theta )\pi_{2}^{{h_{1} f_{2} }} . \) The firm’s optimal period one output, \( q_{1}^{h}, \) is now:

$$ q_{1}^{h} = \frac{{a - c^{h} - (1 - \rho - \theta )\lambda^{h} - I\rho \lambda^{h} - I_{\theta } \theta \lambda^{h} }}{2b} $$

From Eqs. (4a) and (6), it is clear that \( q_{1}^{h} < q_{2}^{h} \quad {\text{if}}\quad p_{ 2} = a - bq_{ 2} , \) hence I = 0. However, if \( p_{ 2} = a - bq_{ 2} - \eta ,\quad q_{1}^{h} > q_{2}^{h} \) is possible, in which case I θ  = 1 (from Eqs. 7, 8), if the negative shock is large enough. More specifically, I θ  = 1 when \( \eta > (2 - \rho )\lambda^{h} . \) Footnote 21 Then,

$$ q_{1}^{h} = \frac{{a - c^{h} - (1 - \rho - \theta )\lambda^{h} - \theta \lambda^{h} }}{2b}\quad {\text{and}}\quad q_{2}^{h} = (a - c^{h} - \eta + \lambda^{h} /2b. $$

We now derive an expression for (f 1). We have \( E\pi (f_{1} ) = \pi_{1}^{f} + E\pi_{2}^{{f_{1} }} , \) with \( E\pi_{2}^{{f_{1} }} = \rho \frac{{(a - c^{f} )^{2} }}{4b} + \theta \frac{{(a - c^{f} - \eta )^{2} }}{4b} + (1 - \rho - \theta )\frac{{(a - c^{f} + \varepsilon )^{2} }}{4b}. \) The expression for \( q_{1}^{f} \) is given by Eq. (8). Using expressions (2b), (8) and (9) as well as the expressions for expected profits, the condition for the firm choosing ‘home’ as its initial location ((h 1) > (f 1)) is nowFootnote 22:

$$ \begin{aligned} \Upphi^{h} < (\rho + \theta )\Upphi^{f} - & \rho \left[ {\frac{{(a - c^{f} )^{2} - (a - c^{h} )^{2} }}{4b}} \right] \\ - &\; \frac{{(a - c^{f} )^{2} - [(a - c^{h} ) - (1 - \rho - \theta )\lambda^{h} - \theta \lambda^{h} ]^{2} }}{4b} \\ - &\; \theta \frac{{(a - c^{f} - \eta )^{2} - (a - c^{h} - \eta + \lambda^{h} )^{2} }}{4b} = \bar{\Upphi }^{h} \\ \end{aligned} $$

We have \( d\bar{\Upphi }^{h} /d\lambda^{h} = \frac{{ - (1 - \rho )(a - c^{h} - (1 - \rho - \theta )\lambda^{h} - \theta \lambda^{h} ) + \theta (a - c^{h} - \eta + \lambda^{h} )}}{2b} < 0 \) (since η > (2 − ρ)λ h). Hence, like in the basic model, we obtain that \( \bar{\Upphi }^{h} \) is lower for higher values of λ h.

About this article

Cite this article

Dewit, G., Görg, H. & Montagna, C. Should I stay or should I go? Foreign direct investment, employment protection and domestic anchorage. Rev World Econ 145, 93–110 (2009).

Download citation

  • Published:

  • Issue Date:

  • DOI:


JEL Classification