5. Foreign direct investment and energy prices

This chapter is a summary of the paper “Do energy prices drive outward FDI? Evidence from a sample of listed firms” (2020[1]) by G. Garsous, T. Koźluk and D. Dlugosch, The Energy Journal, Vol. 41/3. An earlier version of this paper was published as the OECD Working Paper “Foreign Direct Investment and The Pollution Haven Hypothesis: Evidence from Listed Firms” by G. Garsous and T. Koźluk (2017[22]), OECD Economics Department Working Paper, No. 1379.

Foreign direct investment (FDI) has gained more importance with the increasing global dimension of production processes over the past four decades as reported by, for example, UNCTAD. Investing in foreign assets might be beneficial to the investing firm in terms of improving the efficiency of its production lines and gaining access to the local market, while the host country might benefit from technology and know-how transfers as well as from economic development (De Mello, 1997[108]; Saggi, 2002[109]; see also OECD, 2002[110] for a survey). However, improving production patterns and accessing local demand might not be the only reasons for firms to invest abroad – they might want to circumvent tighter domestic environmental policies by shifting their production to countries with laxer regulations. Figure 5.1 shows the share of foreign assets over total assets for the sample of firms underlying this study over the period 1995-2008, along with the evolution of energy prices in the domestic market of these firms. As can be seen from Figure 5.1, both FDI and energy prices both rose from the early 2000s, suggesting that part of the rise in FDI could have been associated with off-shoring in the face of higher energy prices.

The so-called pollution haven effect predicts that firms will respond to tighter domestic environmental policies by shifting production activity towards countries with less stringent environmental policies. Since countries with laxer environmental policies are predicted to gain a competitive advantage in heavily polluting industries (Pethig, 1976[2]; Siebert, 1977[3]; Yohe, 1979[4]), tighter environmental policies should provide incentives for firms to relocate parts of their production processes to countries with laxer regulations (Siebert, 1977[3]; McGuire, 1982[5]; Merrifield, 1988[6]). Assuming that capital is sufficiently mobile across countries, and transportation costs are not too high, this relocation effect is of particular concern for pollution-intensive industries when implementing new environmental regulations. It can lead to carbon leakage, whereby emissions are not reduced by new environmental regulations but simply shifted to other locations, which is a particular concern for global pollutants such as CO2.

The empirical evidence around the pollution haven effect has focused mostly on single country studies until now (see Garsous, Koźluk and Dlugosch (2020[1]) for a detailed review of the literature and Rezza, (2015[7]), for a meta-analysis). This focus is often data-driven, as is probably the geographical focus on the United States. Moreover, the literature is inconclusive so far, with results varying from significant effects (Hanna, 2010[8]; Chung, 2014[9]), to no effects (Eskeland and Harrison, 2003[10]; Kirkpatrick and Shimamoto, 2008[11]; Manderson and Kneller, 2012[12]). Other studies find that the effects vary with, for example, the ability to relocate (Kellenberg, 2009[13]; Cole and Elliott, 2005[14]), the host country’s characteristics (Ben Kheder and Zugravu, 2012[15]) or the firm’s home country characteristics (Dean, Lovely and Wang, 2009[16]). However, the proxies for environmental regulation vary across studies, as does the definition of the FDI variable, which could potentially account for the different effects found in the studies.

This study offers three main contributions to the literature. First, it provides the first cross-country analysis on the relationship between energy prices and FDI at the firm level. Second, the use of energy prices as a proxy for environmental policies, allowing for a better understanding of various environmental policies compared to previous studies, which focuses on pollution abatement costs. Most climate and air pollution policies ultimately affect energy prices – be it directly through additional taxes, or cap-and-trade systems, or indirectly through command-and-control mechanisms – making it a valuable proxy. Third, this study addresses endogeneity concerns about firms’ choices of fuel substitution by using an instrumental variable approach to remove the effects of firms’ substitution choices from the effect of observed energy prices.

The pollution haven effect predicts that it is the difference between the investors’ and the receiving country’s energy prices which drive FDI patterns, rather than changes in the domestic energy prices per se. Therefore, this study uses changes in relative prices as the main explanatory variable. The relative price changes are proxied by taking the difference in domestic energy prices and energy prices in China in the same sector.1 China is the country that had the lowest energy prices over estimation period and was an increasingly important destination country for FDI. In addition, the study also tests whether changes in domestic energy prices directly trigger investments abroad.

There might be industry-level factors which influence both FDI patterns and trends in energy prices, leading to potential endogeneity problems in the estimation of the effects. Technological changes or industry-level shocks to output demand might affect the distribution of industry fuel demand, and consequently energy prices (Linn, 2008[17]). At the same time, these factors might influence FDI decisions of firms, for example, a new clean technology becoming available might make firms rely on a particular fuel more, while at the same time delaying investment in international assets. To avoid such an endogeneity bias, this study follows Linn (2008[17]) and Sato and Dechezleprêtre (2015[18]) by using as an instrumental variable sector-specific energy prices which are weighted according to time-invariant proportions of fuel used at the sector-level. Using these constant weights removes effects of technological change or other industry-specific shocks from the observed energy prices.

This analysis focuses on long-term impacts of changes in energy prices on outward FDI flows by estimating long-differences equations: the first and last observation for each firm are used to estimate the equation below. This implies that short-run effects of variations in energy prices will not be captured in the analysis. The following equation is estimated using a two-stage least squares within estimator:

log(FDI )ijct=α+βlog(EP)jct+γXijct+ μi+λct+εijct

where FDI ijct are the international assets of firm i in sector j in country c at time t. The FDI variable is defined as for-profit assets held by a firm which can be priced. EPjct is, depending on the estimation, either the domestic energy price for sector j, constructed by weighting country-level fuel prices (oil, gas, coal and electricity) by their relative consumption in each country-sector-year, instrumented by fixed-weights energy prices as described above, or the relative energy price difference of country c with regard to Chinese energy prices, instrumented by the difference in fixed-weight prices. Xijct is a set of firm-level control variables including firm size measured as number of employees, total assets and international sales as a measure of international openness. μi is a firm fixed-effect, λct is a country-year dummy which controls country-specific effects such as institutional settings at the country level, and εijct is the idiosyncratic error term.

The dataset is an unbalanced panel from 1995 to 2008 which covers 3 364 firms, operating in 24 OECD economies and 9 different manufacturing sectors. The firm-level data are taken from the Thomson Reuters Worldscope database, which provides balance sheet information about listed firms. Only firms, which are already engaged in FDI are part of the analysis. The data on energy prices are taken from Sato et al. (2019[19]) , who provide a fixed-weight index of energy prices (based on weights from the baseline year 1995).

The main results show that relative prices matter for FDI flows, while there is no statistically significant effect detected for absolute prices alone. Table 5.1. shows the main results of the estimation where the coefficient of the domestic energy price variable is insignificant, while the coefficient of the difference between domestic and Chinese energy prices is significantly positive. An increase of the relative energy price of 1% is estimated to be associated with a 0.5% increase in firms’ international assets. The coefficients of the control variables are also significant and of expected sign, supporting the suitability of the estimated model.

Firms might react to different energy price scenarios in different ways. While an increase in absolute or relative energy prices might incentivise firms to invest more in assets abroad, a decrease in energy prices might not immediately lead firms to withdraw their investments abroad. To study this hypothesis, the energy price variable is interacted with a dummy variable indicating whether the firm faced an increase or a decrease in energy prices over the observation period. The results in Table 5.1. (column 3 and 4) show that relative price changes are significantly positively associated with FDI only in the case of energy price increases, but not statistically significantly so for decreases. A 1% increase in relative energy prices is associated with a 0.7% increase in international assets by firms.

Only a very high carbon tax would be able to influence FDI patterns in an economically significant way. Using a back-of-the-envelope calculation, this study evaluates the effect of the implementation of a high carbon price in developed countries. It assumes that energy prices only rise in countries implementing the carbon tax. A low and a high carbon tax scenario is evaluated. The low carbon tax scenario, based on a carbon tax of USD 15/tonne CO2, would increase energy prices by 5% on average, which would translate into an increase of 0.74 percentage points in the ratio of foreign assets to total assets. The higher carbon tax scenario relies on a carbon tax of USD 55/tonne CO2 and would be expected to raise energy prices by 20% on average for the sample, translating into an increase of 2.6 percentage points in the FDI ratio. As Figure 5.2 shows, these estimated changes in the FDI ratio are rather small compared to the baseline level of the FDI ratio. The calculations have to be taken with a grain of salt, however, as they are based on average effect estimates and rely on a range of other assumptions.

The results of this study show that relative energy prices (i.e. the difference between domestic energy prices and foreign energy prices) matter as a driver of FDI, but the magnitude of the effect is small. The effect is only found for firms facing an energy price increase at home, while a reduction in domestic energy prices is not correlated with a lower amount of international assets. For those firms which did see energy prices rise in their home country, a 1% increase in relative energy prices was associated with an increase of 0.71% in the firms’ international assets. With an average share of international over total assets of 27% in the sample, this average effect on international assets is small.

These results are based on a specific sample of firms, namely firms which are listed on the stock market. While these firms are often larger, thus more engaged in international trade, and contribute a large part in explaining overall changes in the economy, this study is not able to make inferences about smaller companies. However, given that those firms already possess assets across various jurisdictions, they should be in theory the most inclined to undertake foreign investment and international relocation. Therefore, it is reasonable to assume that the effects should be smaller for smaller, less internationally exposed companies not covered in this study. It is also not possible to infer from this analysis whether firms, which did not already engage in FDI at all, will invest abroad because of higher energy prices since this study only covers firms which already engage in FDI.

The simulation undertaken in the context of this analysis reveals that only a high carbon price would have an economically meaningful effect on the share of international assets held by firms. While the analysis relies on a carbon price of USD 55 /tonne of CO2, the OECD has used a carbon price of USD 34/tonne of CO2 as lower bound estimate of the social costs of carbon (OECD, 2015[20]; OECD, 2015[21]). If governments would agree to implement a carbon tax equal to the social costs of carbon or even above, this would, based on the results of this study, only have limited effects in terms of carbon leakage through foreign direct investment.


[15] Ben Kheder, S. and N. Zugravu (2012), “Environmental regulation and French firms location abroad: An economic geography model in an international comparative study”, Ecological Economics, pp. 48-61, https://doi.org/10.1016/j.ecolecon.2011.10.005.

[9] Chung, S. (2014), “Environmental regulation and foreign direct investment: Evidence from South Korea”, Journal of Development Economics, Vol. 108, pp. 222-236, https://doi.org/10.1016/j.jdeveco.2014.01.003.

[14] Cole, M. and R. Elliott (2005), “FDI and the Capital Intensity of “Dirty” Sectors: A Missing Piece of the Pollution Haven Puzzle”, Review of Development Economics, Vol. 9, pp. 530-548, https://doi.org/10.1111/j.1467-9361.2005.00292.x.

[16] Dean, J., M. Lovely and H. Wang (2009), “Are foreign investors attracted to weak environmental regulations? Evaluating the evidence from China”, Journal of Development Economics, Vol. 90, pp. 1-13, https://doi.org/10.1016/j.jdeveco.2008.11.007.

[10] Eskeland, G. and A. Harrison (2003), “Moving to greener pastures? Multinational and the Pollution Haven Hypothesis”, Journal of Development Economics, Vol. 70, pp. 1-23, https://doi.org/10.1016/S0304-3878(02)00084-6.

[22] Garsous, G. and T. Kozluk (2017), “Foreign Direct Investment and The Pollution Haven Hypothesis - Evidence from Listed Firms”, OECD Economics Department Working Papers, Vol. 1379, https://doi.org/10.1787/1e8c0031-en.

[1] Garsous, G., T. Koźluk and D. Dlugosch (2020), “Do energy prices drive outwards FDI? Evidence from a sample of listed firms”, The Energy Journal, Vol. 41/3, https://doi.org/10.5547/01956574.41.3.ggar.

[8] Hanna, R. (2010), “US Environmental Regulation and FDI: Evidence from a Panel of US-Based Multinational Firms”, American Economic Journal: Applied Economics, Vol. 2, pp. 158-189, https://www.aeaweb.org/articles?id=10.1257/app.2.3.158.

[13] Kellenberg, D. (2009), “An empirical investigation of the pollution haven effect with strategic environment and trade policy”, Journal of International Economics, Vol. 78, pp. 242-255, https://doi.org/10.1016/j.jinteco.2009.04.004.

[11] Kirkpatrick, C. and K. Shimamoto (2008), “The effect of environmental regulation on the locational choice of Japanese foreign direct investment”, Applied Economics, Vol. 30, pp. 1399-1409, https://doi.org/10.1080/00036840600794330.

[17] Linn, J. (2008), “Energy Prices and the Adoption of Energy-Saving Technology”, The Economic Journal, Vol. 118, pp. 1986-2012, https://doi.org/10.1111/j.1468-0297.2008.02199.x.

[12] Manderson, E. and R. Kneller (2012), “Environmental Regulations, Outward FDI and Heterogeneous Firms: Are Countries used as Pollution Havens?”, Environmental and Resource Economics, pp. 317-352, https://link.springer.com/article/10.1007/s10640-011-9500-z.

[5] McGuire, M. (1982), “Regulation, factor rewards, and international trade”, Journal of Public Economics, Vol. 17/3, https://www.sciencedirect.com/science/article/pii/004727278290069X (accessed on 17 January 2019).

[6] Merrifield, J. (1988), “The impact of selected abatement strategies on transnational pollution, the terms of trade, and factor rewards: A general equilibrium approach”, Journal of Environmental Economics and Management, Vol. 15, pp. 259-284, https://doi.org/10.1016/0095-0696(88)90002-2.

[21] OECD (2015), Effective Carbon Rates on Energy, OECD Publishing, Paris, https://www.oecd.org/tax/tax-policy/effective-carbon-rates-2018-brochure.pdf.

[20] OECD (2015), Taxing Energy Use 2015: OECD and Selected Partner Economies, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264232334-en.

[2] Pethig, R. (1976), “Pollution, welfare, and environmental policy in the theory of comparative advantage”, Journal of Environmental Economics and Management, Vol. 2, pp. 160-169, https://doi.org/10.1016/0095-0696(76)90031-0.

[7] Rezza, A. (2015), “A meta-analysis of FDI and environmental regulations”, Environment and Development Economics, Vol. 20/2015, pp. 185-208, https://doi.org/10.1017/S1355770X14000114.

[18] Sato, M. and A. Dechezleprêtre (2015), “Asymmetric industrial energy prices and international trade”, Energy Economics, Vol. 52, pp. S130-S141, https://doi.org/10.1016/J.ENECO.2015.08.020.

[19] Sato, M. et al. (2019), “International and sectoral variation in industry energy prices 1995-2015”, Energy Economics, Vol. 78, pp. 235-258.

[3] Siebert, H. (1977), “Environmental quality and the gains from trade”, Kyklos, Vol. 30, pp. 657-673, http://hdl.handle.net/10419/3530.

[4] Yohe, G. (1979), “The backward incidence of pollution control - some comparative statics in general equilibrium”, Journal of Environmental Economics and Management, Vol. 6, pp. 187-198, https://doi.org/10.1016/0095-0696(79)90002-0.


← 1. The energy price data is deflated and converted to constant 2010 USD for tonnes of oil equivalent (toe).

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2021

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.