So after studying the phonologies of many languages, I've noticed the pattern that consonants produced towards the front of the mouth are more likely to be voiced, while those produced towards the back of the mouth are more likely to be unvoiced. For example, it is quite common for a language with a voicing contrast among its stops to lack /p/ or /g/, but it is much rarer to lack /b/ or /k/. Also, in languages that have aspirated stops (phonetically or phonologically), "back" consonants seem to have longer voice onset times, and can therefore be said to be "more voiceless," informally speaking. In English, for example, the VOTs for [ph], [th], and [kh] are around 60, 70, and 80 ms respectively. This pattern can also be seen with non-pulmonic consonants: ejectives, which are unvoiced, are more likely to occur towards the back of the mouth (so that /k'/ and /t'/ are more common are /p'/), while implosives, which are (almost always) voiced, tend to cluster around the front of the mouth (/ɓ/ and /ɗ/ are more common than /ɠ/).
Is there an accepted phonetic/phonological explanation for this pattern? Maybe it's more difficult to produce voiced sounds at the back of the mouth for some reason?