I did look at the section of the joint study that very briefly summarized findings of the CGE modelling done on both sides, some in Japan, and some in Canada. Then they developed a range of estimates, like other CGE models of a trade agreement, based on the assumption of full employment and the maintenance of full employment. In essence, this forces a positive result, because when you make that assumption, everyone in the country is going to be employed after a free trade agreement, and market forces will ensure that they're doing something relatively more productive, just as is assumed in the traditional comparative advantage model of free trade.
The problem, of course, is that the full employment assumption and the representative household assumption, which also assumes that every Canadian will share equally in the gains and losses from a free trade agreement, are completely invalid in the real world economy. What really happens with a trade agreement in the real world is that if your net exports increase, if you sell more than you import, then you will get benefits, not just from mutual specialization, but from more output and more employment. If the reverse happens, then you'll get net losses. So either can happen. It depends on the competitiveness of your country's products, and the relative impacts of the trade liberalization on the two sides.
The conclusion of that joint report with the simulations of the CGE models, which were not fully documented there.... I would have to see more detail on how the models were actually specified, but I do not accept them as credible. In fact, I have written a scholarly article, which I've referred to your clerk, detailing the way CGE models are completely dependent on those unrealistic assumptions and proposing what I would view as more relevant and more reliable methodologies for estimating the impact of trade liberalization on employment, income, and investment.