A relatively new approach for evaluative research is 'Realistic Evaluation'. See the works of Pawson and Tilley. Their work is underpinned by a critical realist philosophy.  

If you are looking to evaluate the impact of tourism policy on the economy of a country, you can consider CGE (Computable General Equilibrium) modelling and DSGE (Dynamic stochastic general equilibrium modelling) modelling. Although there are bit complex methods, those provide very useful information for policy making. You can use Mathlab software for analysing data.

Hope this information is useful to you.

Kind regards

Thushara

it depending on your developing study.  If your study is focusing on tourism behavior study, the growing methodology to day is applied structural equation modeling SEM based on scale measurement. please consider SmartPls 3 developed by Prof Ringle and Sven Wende of Humburg Insitute Germany. It many advantageous tools you can take benefits to developed information quality need in tourism study.   

Hi, Adenike, we have experience in Bali tourism growth and it impact to generating income for society. However, tourism fail to make income redistribution to local community. we recommend you to start with community participation ( see Tosun, 2000) and also Timothy and Tosun (2003) for enhance local community participation in tourism. You can set back to local governent policy as instrument for tourism planning and it contrubution to support sustainable tourism based on community participation, Local communuity might be the key to income redistribution and sustainable tourism might be  achieves. 

Essentially, your methodology is the theory of how your research will proceed; so it will depend on what aspect(s) of tourism policy that you want to evaluate and/or assess; and therefore will be primarily quantitative or qualitative.

It all depends how do you define tourism policy.

Anyway I would start with monitoring of tourism numbers (expenditure, arrivals etc.). 

I think a wiser move here would be trying to take up your own parameters that would be aptly suited to the destination/region for which the policy has been drafted.

As has been previously stated, the main question would be on what aspects would would like to weigh it? Once the same has been decided, your approach (QUAL, QUAN, Mixed could be formulated) and then you would be in a better position to use a combination of models or just one if you would prefer.