You can only ever decrease the data rate with no loss of quality if you use what is known as a lossless codec (Such as HuffYUV).
Your question should really be "how do I minimise quality loss"? That yields so many variables it's impossible to answer. It depends on the dimensions of the video, what's happening in it, and what compression it's already been subjected to.
Your best bet is to decide out what file size you want it to be in MB, divide that by the duration in seconds then multiply that by 8388 to get the data rate in kbps
So for a 2½ minute vid that we want to fit into 10 MB the maths is:
10 ÷ 150 x 8388 = 559.2 kbps
Decide the smallest pixel dimension you think you can get away with (e.g. 355x288) and recode it using that size and the calculated bitrate using a variety of codecs. Choose the best looking one from the various outputs. It may change from one piece of footage to another even iff all the other variables (size, duration, bitrate) stay the same.