The current research and analysis of the wages data could be considered 
adequate in order to implement a model on a website to allow citizens to 
see if they are being paid a fair wage, but further examination and 
investigation could be done.

More specifically, it would be appropriate to enlarge the number of 
observations in order to represent a more appropriate randomized sample of 
the general population.  In the same respect, perhaps other variables could 
be added and tested for their significance in the model, such as 
generalizing South to a Location factor (North, East, Mid-West, etc.). 
More Occupation and Sector groupings could also be added, such as 
Education, or Medical.

Starting with all the two-way interactions between all of the covariates 
and then working backwards to find the underlying model would be more 
applicable than just considering individual covariates.  The correlation 
between any of the covariates should also be considered, and should be 
handled accordingly.  Lastly, any obvious outliers could be removed from 
the dataset in order more accurately represent the "typical" working 
citizen.

For any new models found through this added research and analysis, it would 
be beneficial to produce conditional effects plots in order to map out how 
the covariates affect not only each other but also the model in general. 
In the end though, the result from the multiple regression could be used in 
order to calculate an "expected" wage for a citizen who enters their 
specific covariate information via the website.  Calculated 95% confidence 
intervals for the model would then allow the citizen to examine whether or 
not their expected wage falls within the "fair" wage range.