Measuring trade policy effects

I was going through the Results Framework Document (What's RFD) for Ministry of Commerce here. In short, the RFD seeks to address three basic questions:
(a) What are ministry’s/department’s main objectives for the year? (b) What actions are proposed by the department to achieve these objectives? (c) How would someone know at the end of the year the degree of progress made in implementing these actions? That is, what are the relevant success indicators and their targets which can be monitored?

The primary objective as per the RFD for Commerce ministry is to "To provide policy support for increasing India’s annual export growth." 

The item under action points for this goal reads, "Action-1: An export level of US$ 300 billion is targeted to be reached at the end of 2011-12". And

"Action-2: Disposal of Appeals overdue under the Foreign Trade (Development & Regulation) Act, 1992."

I will focus only on Action 1 as Action 2 is more of a service obligation that must be met.

Action 1 has these details: 

Success indicator: "Degree of achievement of the target set for the year in US $ Billion" and the target/criteria says "300 Billion $ for excellent, 270 for Very Good, 240 for good" and so on. 

The issue that bothers me here is this:

Can we really attribute the trade performance to trade policy like this? Do we really think that the fiscal incentives and other measures directly lead to increase in exports in such a direct manner? Do we have methods though which we can say that if I provide 'X' amount of incentive to a particular sector, this would generate an export increase of 'Y%'. One might say, it is understood that that's how it works,  and it cannot be easily measured. In that case, my further question would be, how do you decide as to how to split the Big X, i.e. the total incentive spending into various small Xs for different sectors? 

One of the ways of doing it could be, to have a marginal export improvement measure (wrt. incentives), for all sectors and then distribute the X based on this measure. Or the distribution could be also based on other policy goals, such as saving a sensitive (employment/livelihood etc) industry or sector. However, the cause and effect should be linked and THAT should be measurable/observable. 

To summarize all the points above, the questions are: Do we have models that can build a case to provide policy interventions? Can we simulate the effects of policy changes and study the results on the export performance, and the economy as a whole? 

A good place to start would be to have comprehensive and timely data on trade performance, to study the effects of a new policy that has been implemented. When the data itself comes with a lag of more than 6 months, how reasonable it would be to say in the annual RFD, that policy implemented has led to export performance of 300 billion dollars?