was mentioned earlier, Chile applied a uniform 11 percent import tariff to imports of all
products from all countries, which fell to 7.5 percent by 2005, the last year of our sample.
The Canadian simple average tariff for agricultural industries was 7 percent in 1997, while the
applied import tariff weighted by trade shares was 11.2 percent that declined to 4.6 percent
in 2005. By 2005, 65 percent of all HS6 categories of Chilean exports were exported duty-free
into Canada, while only 35 percent of Canadian HS6 categories received duty-free treatment
in Chile. As a result of this prolonged liberalization of the Chilean agricultural sector, the
effect of the CCFTA may not be completely reflected in the data currently available.
If a different time path for tariff reductions is the only reason for observed asymmetry
in the effect of the CCFTA on bilateral trade, then the elasticity of Canadian imports and
exports with respect to the FTA tariff preferences should not be different. To verify if
this is the case, we used an econometric specification that comes from a reduced form of
microeconomic partial equilibrium model (Clausing, 2001). The basic empirical specification
takes the following form:
∆ ln(Xct) = β0 + β1∆tariffct + βT [YearEffects]+εct (3)
where c denotes industry and t denotes time. Xct is the value of Canadian industry c
imports from (exports to) Chile at year t, tarif fct is the Canadian (Chilean) ad-valorem free
tariff for the industry c imports from Canada (exports to Chile) at year t,and∆ denotes
the one year time-differencing.
Table 6 presents estimation results for specification (3). The GLS estimates in columns
(5)-(6) show no effect of Chilean tariff preferences granted to Canadian exports to Chile: the
16