Javorcik (2004)

Question 1

What was Javorcik (2004) about?

What are the findings of the paper?

Answer 1

Many countries strive to attract foreign direct investment (FDI) hoping that knowledge brought by multinationals will spillover to domestic industries and increase their productivity.

In contrast with earlier literature that failed to find positive intra-industry spillovers from FDI, this study focuses on effects operating across industries.

The analysis, based on firm-level data from Lithuania, produces evidence consistent with positive productivity spillovers from FDI taking place through contacts between foreign affiliates and their local suppliers in upstream sectors.

The data indicate that spillovers are associated with projects with shared domestic and foreign ownership but not with fully owned foreign investments.

Spillovers driven by partially foreign owned affiliates. Speculates fully foreign owned affiliates do not create spillovers because do not source inputs from domestic suppliers

Question 2

What are the limitations of Javorcik (2004) ?

Answer 2

External validity on the grounds that Jarvocik(2004) is only based on Lithuania in the immediate era preceding Soviet Union’s disintegration.

Question 3

What was the OLS strategy for Javorcik (2004)?

What are the limitations of OLS?

Answer 3

OLS Strategy:

Run regression firm level output on Horizonal, backwards and forwards variable controlling for firm’s capital, labour and material input, as well as controlling for foreign share of firm’s ownership

OLS Findings:

1. Foreign owned firms are more productive than domestic owned firms
2. OLS: both upstream and downstream spillovers observed significantly (lagged and unlagged) [whereas Olley-Pakes method only found upstream spillover]

OLS limitations:

1. OVB: could be factors that are both correlated with firms’ productivity and their exposure for FDI firms
2. Dependent variable is firm-level, whereas independent variables is industry-level. This leads to downward bias in standard errors
3. Endogeneity bias at firm level – firms’ input choices are correlated with unobserved productivity

Question 4

What is the Olley-Pakes method used by Javorcik (2004)?

Answer 4

‘Olley-Pakes methods’: is a way to try to obtain estimates of firm’s productivity, accounting for the amount of factor inputs the firm uses, and accounting for the fact that those inputs are chosen endogenously given firms’ unobserved productivity.

Question 5

How does Javorcik measure the potential for within-industry, upstream and downstream spillovers from FDI? Your answer should explain how these variables are constructed. (4 marks)

Answer 5

1. Within-industry spillover measure: Share of industry output produced by foreign owned firms. (Weighted average of foreign ownership share of firms in the industry. Weights: firm output. [2] )
2. Upstream spillover measure: Share of industry output sold to foreign owned firms (Weighted sum of within-industry spillover measures (of all sectors but your own). Weights: Share of output supplied to each downstream sector. [2] )
3. Downstream spillover measure: Share of industry inputs purchased from foreign owned firms

Constructing independent variables:
Within Industry spillovers:

Share of industry output produced by foreign owned firms

Upstream spillovers (Backwards):

Share of industry output sold to foreign owned firms

Downstream Spillovers: (Forwards)

Share of inputs purchased from foreign owned firms

Regression:

Run regression firm level output on Horizonal, backwards and forwards variable controlling for firm’s capital, labour and material input, as well as controlling for foreign share of firm’s ownership

Question 6

To test for FDI spillovers, Javorcik regresses firm-level output on the firm’s inputs of capital, labour and materials, the share of the firm’s equity owned by foreign investors, variables measuring the potential for within-industry, upstream and downstream spillovers in the firm’s industry and year, region and industry fixed effects. Why might OLS estimation of this specification lead to biased estimates of FDI spillovers? How does Javorcik deal with these concerns? Does Javorcik’s estimation strategy eliminate all potential sources of bias? (5 marks)

Answer 6

Why might OLS estimation of this specification lead to biased estimates of FDI spillovers? [2]

• (i) OVB: Spillovers and unobservables

(ii) Correlation of input choices with unobserved productivity.

How does Javorcik deal with these concerns? [2]

• For (i): Estimation in first differences to kill off constant unobservables (remember panel data).
• For (i): Also controls for industry concentration and level of downstream demand.
• For (ii): Try to estimate firm’s productivity directly

Does Javorciks estimation strategy eliminate all potential sources of bias? [2]

No. Remaining bias: Allocation of FDI across industries not random.

Question 7

Does Javorcik find evidence of FDI spillovers?

Suggest a possible explanation for the pattern of spillovers Javorcik finds. (4 marks)

Answer 7

Does Javorcik find evidence of FDI spillovers? [1]

Under Olley-Pakes method: Upstream: yes. Within or downstream: no. (OLS Finds both upstream and with-in industry spillover)

Suggest a possible explanation for the pattern of spillovers Javorcik finds. [3]

• Relationship learning.
• FDI effect: MNEs may have incentive to transfer technology to suppliers.