Trends in free trade in TPP, RCEP, and bilateral FTA suggest that it is unrealistic to think that high tariffs will remain on agricultural products in perpetuity. The standard economic method of trial calculation of FTA effects is the use of applied generable equilibrium analysis through the use of the GTAP model. However, the elasticity of substitution among import products set in GTAP adopts the so-called "Rule of Two," whereby values are double those for domestic products and import products; the econometric evidence is, therefore, unclear. In simulations of tariff elimination through the GTAP model and other schema, the elasticity of substitution value plays an important role, so it must be accurately measured. One major issue surrounding the calculation of elasticity of substitution is simultaneity. Theoretical research into the above begins with Leamer (1981), with Feenstra (1994) using the features of panel data to derive an empirical methodology, and later research by Soderbery (2015) describes a methodology utilizing Limited Information Maximum Likelihood (LIML) and constrained LIML. In this paper, we focus on the Japanese dairy products market as an example of a market in which high tariffs persist. By utilizing Soderbery’s (2015) methodology, (1) we estimate the elasticity of substitution of dairy products among Japanese imports and conduct a simulation of free trade comparing the elasticity of substitution defined respectively in the GTAP model and in the parameters estimated in our research, (2) with the respective results compared to consider the difference between divergent elasticity of substitution. The results show that the estimates of elasticity of import substitution for dairy products in Japan yield a lower value than those set in GTAP, and our findings quantitatively suggest that, as a result of the above, FTA simulations produced a smaller impact in terms of trade diversion effects.