Determinants of foreign direct investment across China

Abstract

We analyze the spatial and temporal variation in foreign direct investment (FDI) among China's 30 provinces from 1986 to 1998. Motivated by Naughton (Brooklings Pap Econo Activ 2 (1996) 273), we distinguish our study from similar studies by examining changes in importance of FDI determinants through time. We do find supporting evidence. This is due to the shift in the nature of FDI in China. We also find that the cumulative FDI relative to cumulative domestic investment has a negative impact on the new FDI. Provincial officials have to improve the investment environment. Otherwise, multinational corporations may choose to invest in provinces with fewer FDI competitors. Our analysis is robust across different specifications. However, it explains the FDI distribution in the coastal provinces better than it does for Central and Western provinces.

Introduction

Interest in the study of foreign direct investment (FDI) has grown rapidly. Existing studies of FDI in the US help us to understand factors that are important to attracting foreign investments across different states.¹ In this paper, we examine if these factors are also important for FDI distribution across different provinces in a developing country, China. For sure, there are many studies of FDI on developing countries but most of them are cross-country analyses.² As such, the interwoven relationships between social, cultural, economic, and political factors are difficult to delineate. By focusing on only one country, we can make a cleaner study on the economic determining factors that attract FDI.

Analysis of FDI in China is also timely. After China's successful talks with the US and the European Union on WTO and US granting China PNTR, China entering into the WTO is almost a sure thing. China's entering of the WTO would likely sparkle another round of FDI projects. MNCs in Shanghai and Tianjin have already planned to expand their investment scale in China. Erickson plans to double its current investment amount of US$3 billion by 2001. On the other hand, China has promised to open more of its servicing industries to foreign investments, especially on areas like financial sector, insurance, telecommunication, trade, and tourism. In fact, China has already been the second largest host of FDI in the world since 1994, next only to the US. An American credit-rating agency projects that by 2005, annual FDI in China will reach U$100 billion.³

However, there are only a few empirical studies on the overall FDI situation in China. Wang and Swain (1995) examine the host country determinants of FDI in China. They find that the FDI in manufacturing sector is positively related to China's GDP, GDP growth, wages, and trade barriers, but negatively related to interest rate and exchange rate for the period of 1978–1992. Chen et al. (1995) study the effect of FDI on China's output and find that the FDI has a positive impact on the output growth between 1978 and 1990. Using cross-country data, Wei, 1995a, Wei, 1995b find that despite the large amount of FDI China has received in recent years, the country still appears to host too little FDI compared to an ‘average’ host country.

The present analysis focuses upon the spatial and temporal variation in FDI among China's all 30 provinces from 1986 to 1998. Such studies are minimal except a few that use data at the city level with a relatively short time span. Heid and Ries (1996) study 931 joint ventures in 54 cities from 1984 to 1991. They intentionally exclude investments by overseas Chinese (Hong Kong, Macau, Singapore) which probably have a different set of location determinants due to familial, linguistic, and cultural ties. Their conditional logit regression shows that cities with good infrastructure, established industrial base and foreign investment presence are more attractive to investors. Wei (1995a) also looks at individual cities. He finds clear evidence that in the late 1980s, FDIs contribute to higher growth of the cities.

Kinoshita (1997) examines the data from a special survey conducted by the World Bank in 1992 on 468 firms in eight cities in China (six located in coastal provinces and two in inland cities). She investigates the possible effects of FDI on improving a firm's total factor productivity during the 1990–1992 period. She finds no evidence that foreign investment helps increase the productivity growth of local firms via foreign joint ventures, foreign linkages, and the mere presence of foreign firms in the industry. Hence, she concludes that opening up to foreign investments is not sufficient for a country to benefit from foreign technology spillovers.

Branstetter and Feenstra (1999) have an interesting working paper, which does look at FDI in China at the provincial level. However, their focus is different. They use 29 provincial data over the years 1984–1995 to estimate the structural parameters of government's welfare function. They want to examine the tradeoff between the benefits of increased trade and FDI against the losses incurred by state-owned enterprises as a result of such liberalization. Indeed, they find that the government places much heavier weights on the output of the state-owned enterprises than on consumer welfare, although such preference has declined somewhat over time. They hence post skepticism on China's entering into the WTO. Similar to our study is a paper by Cheng and Kwan (1999). They look at data from 29 provinces from 1986 to 1995 and observe agglomeration effects of foreign capital stock.

However, there is an important fact that none of these studies has taken account into, and that is the big differences in the scale and nature of FDIs in the 1980s and the 1990s. In his important paper, Naughton (1996) points out the critical role that Hong Kong and Taiwan has played on the FDI in Guangdong and Fujian provinces, especially in the 1980s and early 1990s. The total FDI never exceeded 1 percent of GDP before 1991 with over 40 percent of all FDI in Guangdong and 10 percent in Fujian. This is probably due to the linguistic and cultural ties of Guangdong with Hong Kong and Fujian with Taiwan. In fact, between 1984 and 1990, Hong Kong accounted over 50 percent of China's total FDI all along (Wei, 1995a, Wei, 1995b). Naughton (1996) also points out that until 1991 virtually all of the industrial output of FDI was exported. Such lop-sided phenomenon began to fade away when China began to offer significant domestic market access to foreign investors in 1992. The output of these foreign-invested enterprises (FIEs) has increasingly been directed toward the domestic market. Given this, it is conceivable that the factors drawing FDI into China in the early period may not be the same, or at least may not be as important as those in the later period. Pooling everything together under one regression model hence ignores the possible dynamic complexity of the issue.

As an attempt to address Naughton's point, our paper makes contributions on this line of research by examining FDI determinants over different periods of time. Furthermore, we contrast results based on the full sample of 30 provinces against those based on 28 provinces, excluding Guangdong and Fujian. By doing so, we are able to see if FDI determinants change through time and if Guangdong and Fujian, two of the earliest provinces opening to and drawing in most of the FDI, have different attributes attracting FDI.

Our results do show that the importance of the FDI determinants changes through time. However, there is no evidence that Guangdong and Fujian are very different from other provinces in terms of attracting FDI. Surprisingly, we find that the cumulative foreign investment relative to the cumulative domestic investment has a negative impact on the new FDI. It provides implications to both policy-makers in China and foreign investors interested in the China market. Specifically, provincial officials have to improve the investment environment. Otherwise, multinational corporations (MNCs) may choose to invest in provinces with fewer FDI competitors.

The analysis is organized as follows. Section 2 gives an overview of FDI development in China. Section 3 discusses the conceptual framework. Data and empirical methodology are described in Section 4. Section 5 presents the empirical results, and Section 6 concludes the paper.