![Sample our Economics, Finance,Business & Industry journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5931/TOC-Subject-Sample-2021-Economics-Finance-Business-Industry.jpg"})
Abstract
This paper quantifies the role of FDI in the remarkable fivefold gains in eco-efficiency in China’s manufacturing during the period 1995–2006. We refine the conventional index decomposition method and apply it to China’s emission. We find that (i) foreign-invested enterprises (FIEs) have shifted production toward cleaner industries; (ii) the better abatement technique of FIEs explains 23% of the eco-efficiency gains; and (iii) another 69% of the eco-efficiency gains come from by progress in domestic firms’ abatement technique, which is associated with forward environmental spillovers from FIEs. Findings (ii) and (iii) lend support to the so-called Pollution Halo Hypothesis.
Notes
* Accepted during 2014 APJAE Symposium in Kokura, Japan, Taipei office.
No potential conflict of interest was reported by the authors.
1 See, for instance, “As China Roars, Pollution Reaches Deadly Extreme,” New York Times, 26 August 2007.
2 Emission and output data are taken from various issues of China Environmental Yearbook. Recent studies, including bib35 and bib28, have begun to recognize that China is decoupling pollution with economic growth. bib13 also find that, despite China’s rapid growth in exports, emission of air pollutants has either fallen or remained roughly constant.
3 For example, bib2 quantifies the three effects in the environmental consequences of trade liberalization; bib27 quantifies the three effects in the cleanup in US manufacturing.
4 See bib38 for a detailed discussion.
5 According to proponents of the Pollution Halo Hypothesis, multinational firms often refrain from lowering their environmental standards in their affiliates in developing countries to reduce (i) liability risks (bib29), (ii) transaction costs (bib39), and (iii) reputation costs (bib9). Other supporting evidence includes the case studies in bib17, bib37, and bib31, and the empirical studies in bib15, bib33, and bib8.
6 See bib16 and bib1 for supporting anecdotal and empirical evidence.
7 Evidence for the Pollution Haven Hypothesis is limited (see, e.g. bib7; bib15; bib36). bib11 provides a comprehensive literature review. bib14 find pollution-haven-seeking behavior in FIEs in China that are from Hong Kong, Macau, and Taiwan, but not in FIEs from OECD countries.
8 In Chinese statistics, FIEs are referred to entirely foreign-owned establishments and Chinese-foreign joint ventures.
9 Data sources are provided in details in Section .
10 The annual sample of CEY covers all the polluting facilities registered at the “county level and above.” CEY reports the industry-level data on pollution and output of the sampled facilities every year before 2007. bib13 and the data-set of Pollution Indicators and Status for China by the World Bank also refer to the CEY data for calculating China’s industrial emission intensities.
11 Literature suggests that FIEs operating in developed countries may adopt the same environmental as in their home countries because of the concerns about liability risk, transaction costs, and reputation costs. For example, bib29 argues that the publicity of poor environmental performance in foreign affiliates will raise multinationals financial risks; bib39 suggests that multinationals aim to avoid transaction costs and smooth streamline international trade rather than reap residual savings of degrading host country’s environment; bib9 argues that multinationals have large export markets in the developed world where consumers are environmentally aware. Although China’s FIEs might adopt inferior production or abatement techniques relative to those adopt in FDI source countries, it is also true that only a small fraction of firms in a country become multinationals which are on average are more productive than other firms (bib22). As productivity is negatively associated with pollution intensity, using the source country’s average pollution intensity to proxy for the FIEs’ pollution intensity may not always lead to an underestimate of FIEs pollution intensity.
12 Regressing on
for various economies
, the coefficient
is in general close to 1. In particular,
,
,
,
, and
, and all are significant at 1% level. Data for industrial output shares,
, are obtained from the World Input–Output Database, except that data for
are obtained from the Singapore Economic Development Board.
13 The lagged terms are taken up to four years in our empirical analysis. If greater lags are taken, the estimated FIEs’ pollution intensity of some industries turns out to be even larger than China’s average pollution intensity, which does not seem to be realistic.
14 “Others” includes Taiwan, Hong Kong, and free ports. Since Hong Kong and free ports have only a small amount of manufacturing, and their local pollution intensities may not be representative of their Chinese affiliates’ pollution intensities, we use Taiwan’s pollution intensities as proxies for those of all the FIEs originating from the “others” group.
15 Decomposition results for alternative estimates by and
appear to be similar to Figure (b), and are documented in Table .
16 Supporting evidence for environmental spillovers from FIEs to domestic firms is found by bib1. Using the Argentinean firm-level survey data, they find that the number of environmental management systems among the eight possible types reportedly implemented by domestic firms, which also export, increases with the presence of FIEs in the downstream industries. Our study complements to theirs by looking into the environmental performance (emission intensities) of domestic firms and the possible environmental spillovers from FIEs.
17 see bib6 provided a real-life example where Ford, Siemens, and Toyota required that their suppliers to adopt the International Standards Organization 14,000 of environmental standards.
18 Since firm-level data for abatement techniques are unavailable, we have to rely on industry-level data (with a breakdown between FIEs and domestic firms) for our regressions.
19 While this paper is interested in horizontal spillovers in production or abatement techniques, many studies examine the horizontal spillovers in total factor productivity (see, e.g. bib4; bib20; bib25; bib32).
20 Data for and
are constructed using the Input–Output Tables of China 2002. Since the Input–Output Tables are available only in selected years, we follow the literature to assume that
and
are constant over time.
21 We construct the backward-linkage and forward-linkage variables following bib24 and bib1.
22 This percentage share is calculated by dividing China’s exports of manufactured goods to developed economies and high-income developing economies by its total exports of manufactured goods for the period 1995–2006 and then taking the annual average. Exports data are obtained from the UNCTAD Database. Some literature shows the export propensity affect an enterprise’s environmental performance is bib5 find that a higher export intensity is associated with more environmental patents. bib10 find that a higher export intensity is associated with a lower intensity of carbon dioxide emission.
23 see bib18 argues that an increase in demand for innovations in abatement techniques, which promotes entry into and competition in the related research and development activities, helps lower the costs of abatement equipment. In that case, a higher demand for such innovations will help reduce the pollution intensity.