Empirical Model:
Prior empirical works on the impact of export credit programs on US export include Koo
and Karemera (1991) and Diersen (1995).17 Other empirical studies analyze import
demand models on the impact of US non-price export promotion programs on US exports
(Le et al (1998); Halliburton and Henneberry (1995)).18 Our empirical import demand
model is based on a general consumer demand model with the inclusion of demand
shifters such as exchange rates and a variable of cost savings resulting from export credit
to test our concept of additionality. Thus an importing countries’ demand for wheat is a
function of own price, price of substitutes, income and domestic supply. Thus our model
is specified as follows:
P
Mi= f(jPc,GDPi,DOMi,EXi,PVCi) (1)
Common demand theory applied to import demand model states that imports of an
importing country i from a exporter j (Mij ) are a function of the price of the exporting
country (Pj) and price of its competitors (Pc) .19 Exchange rates (EXij) determine the price
17 Koo and Karemera (1991) used a similar approach to analyze the impact of export credit programs.
Using a gravity model approach, they incorporate dummy variables based on in the periods in which export
credit under GSM-102 was offered to capture the shift in the import demand. On the other hand Diesren
(1995) on the other hand tested for additionality under the GSM-102 program using an intertemporal
consumption model and “loans” as choice variables . Our study differs from Diesren (1995) in that we
include the net present value of cost savings into the budget constraint Empirically, we calculate the
present value of cost savings based on semi-annual payment and use actual repayment period while
Diersen (1995) assumed single payment and fixed repayment period of 3 years.
18 The estimated coefficient of the amount spent on non-price export promotion programs in these studies
are interpreted as additionality, that is, additional imports due to the increased spending on these programs.
19 See Lord, 1991 for details on application of common demand theory to trade models. Ideally we should
use the Canadian price of wheat and prices of other competetitors’ , which compete directly with the US
in the international wheat market as a substitute for US wheat. But due to lack of data we use the ROW
price. Moreover as the US price and the ROW price move closely together we use a price ratio to avoid
multicollinearity problems.
24