The first chapter takes an overview of the Chinese steel industry. The intermediate

goods trade has become a signicant feature of the international economy. Nevertheless,

questions regarding price negotiation and the determinants of importing

firms' profitability remain unanswered. Using firm level data, we attempt to address

these issues in the context of the Chinese steel industry. Despite being the largest

producer of steel in the world, the Chinese steel industry has maintained a very

low level of profitability. This paper suggests that the low profitability of Chinese

steel producers results from the abnormally high degree of market segmentation in

China. A recently developed econometric method in panel data spatial analysis

is adopted here, to explain the level of geographic fragmentation in the Chinese

steel industry. Our results reveal that local steel production depends only on local

demand rather than on cross-region demand. Production is responsive, as a 10%

increase in local GDP induces more than 8% increase in local steel production, while

the cross-province spill-over demand is insignificant under several reasonable model

settings. Less efficient firms survive because of the segmented market, so Chinese

steel producers realize lower profit in the face of high input prices.

The second chapter develops a model of global supply chain to study how profits

are shared between intermediate input suppliers and final good producers. Differences

in market structures are shown to be main driving forces in profitability

differences along the global supply chain. Applying Melitz and Ottaviano (2008)'s

framework of heterogeneous firms into the problem, we model the downstream (final

good) market as the monopolistic competition, while the upstream (intermediate input)

market in oligopolistic (Cournot) competition. We show how increases in the

entry cost in the upstream market and segmentations in the final good market increases

(decreases) the market power of intermediate input (final good) producers,

which increases (decreases) the profitability of intermediate input (final good) producers.

We also show how increases in the demand of the final good affect the price

of the intermediate input, which determines the profit sharing between intermediate

input and final good producers. Using firm level data from the Chinese steel

industry, we calibrate the model. Our results show a 20% increase in the final good

demand would induce 25% increase in the input price. Our results also suggest

the regional trade costs in the Chinese steel market are about three times as firm's

average marginal cost, and a 10% decrease in regional trade costs would lower the

input prices by 22%.

The third chapter estimates a dynamic structural model of rm investment and

importing decisions. Using firm level data from the Chinese steel industry, both

activities are found to have positive effects on productivity. Particularly, firms

engaging into both activities enjoy a 3.72% productivity premium in the long run.

Moreover, the result suggest that in the Chinese steel industry there is a huge entry

cost and fixed cost in the import market for raw materials. These costs create a self-selection

issue in the import market. After controlling for this issue, I find that more

productive firms benefit to a larger degree from importing. Furthermore, simulation

of the import market shows that a 10% decrease in entry and fixed costs would net

8.9% productivity gains for a typical firm.

Evidences presented in this dissertation shed lights on many aspects. First, it

firstly documents the market structure in the Chinese steel industry, using a newly

developed panel data spatial analysis approach to reveal a segmented market in this

industry; Second, we introduce the market vertical linkage to a classical new trade

model to show the relationship between the inter-regional trade barriers and the

upstream market price, which is a valuable contribution to the international trade

literature under the global value chain context; Last but not the least, firms' importing

and investment behavior as well as their effects on productivity dynamics

are structurally estimated, so that the gains from import is well measured while

addressing the endogeneity issues. In all, my dissertation focuses on the low profitability

issue in the Chinese steel industry. Based on detailed firm level data, it

provides reasonable explanations from several perspectives, which also shows many

potentials for researches in the future.