Help needed (Econometric/stat emp. problem)

[economicus]

Hi, knowing that here are some very good econometricians/statisticians I would like to ask for your help.

The problem is that I have results of several surveys that are as follows:

1. How big is a problem of financing you business?
Country A Country B Country ....
small: 5% 10%
medium: 15% 15%
big: 60% 35%
huge: 20% 40%

2. How big is a problem of .........?
....

Now I need to include these characteristics into regression equations, such that I need to somewhat transform the responses into a single index of the problem severity. I was thinking of taking only the % of respondends who answered "huge", but as it can be seen this would imply that country B has more severe problems... but country A has a much greater share of respondents who answered "big". Then I thought of taking a sum of both shares (big and huge) or some weighted average of those. Is there perhaps any better statistical or econometric technique to solve this puzzle? Thanks a lot!