A search on Google Trends for the phrase "free blog" delivers the following distributions:
These are normalized results, which means that they indicate the proportion of searches from those areas which mention the phrase. The results are then ranked. So the first chart above indicates that Manila has the highest percentage of searches for "free blog" compared to all searches from that location.
Of crouse, these results are not totally reliable as "free blog" is not in the native languages of those regions. However, young people in Malaysia and Singapore certainly do use English and have education systems which promote it.
Results from Google Trends for 'start blog' (not the phrase, but the 2 terms) gives a bias to US, English language searches. However, regionally, the top 3 are Singapore, United States and India.
It is a little hard to be comfortable with the results from Google Trend as there isn't really enough information describing exactly what it does. For example - why does the trend for 'start blog' look like this:
My guess is that, due to the sampling a search expression either makes it into the database or not. 'Start blog' passed the threshold required to make it in early 2005 (note that since then it never dips below that first data point).
What we really want to do is trend the number of blogs by region...