Russia and Middle East) and finally, as a complement, also “Home” has been considered
so not to exclude from the sample the counterfactual.
From this, we construct localization-specific versions of the indexes presented in section
3.2, respectively on import and export practices separately and then on their combination.
For firm i at time t we have the share of intermediate inputs imported from each origin and
the share of export to each destination calculated as,
iX
Imported Interm.it *
Mat. Inputs it j
(share of Imports from X )it =
(Imported Interm. from X)it
Mat. Inputs it j
(9)
and
eX it = ( Export⅛ Ï * (share of to X ) = ( (Eχports to X)it Ï
(10)
it к Salesit Jv m к Salesit J
where X represents “North”, “South”; “Asia” and “Home”.
From this we then construct the localization-specific version of (5), for firm i at time t , is
defined as:
where j= “North”, “South”, “Asia” and “Home” and also i = “North”, “South”, “Asia”
and “Home”. Therefore, j indicates the origin of import, and i the destination of sales.
This indeχ takes the value zero if the firm does not have contact with any of the two areas
considered. While it takes the value one, its upper bound, if a firm imports all of his
material inputs from the same area and sells all its output to the same area. Many firms in
our sample are not eχclusively dealing with one single geographical area and this indeχ
accounts for all the trade flows of each firms. If for eχample a firm buys 30 percent of its
IjEiit
Imported Intermediates from j it
—
Material Inputs it
* Exports to iit
Sales.,
к it J
(11)
23