Alyosha We don't have any evidence from the Key Stage 2 SATS, because their format was changed shortly after the first phonics-taught cohort began taking them. There is evidence from the Key Stage 1 SATS, however.
The first cohort to be taught via systematic phonics (or more correctly, who were supposed to be taught that way) started school in 2007, which means the first group taught via phonics took their SATS in 2010. These are the relevant SATS results:
Number of children achieving the expected level of reading comprehension at the end of Year 2
2007-2009 – 84%
2010-2011 – 85%
2012 - 87%
2013 - 89%
2014 - 90%:
Note that this gradual rise began with the cohort who were in reception for the academic year 2007-08, that is, the first year that phonics teaching was introduced.
Number of children with SEN achieving the expected level of reading comprehension at the end of Year 2 (figures are not available for before 2011).
2011 - 52%
2012 - 55%
2013 - 58%
2014 - 60%:
For SEN children, that's a 15% increase in 3 years.
One key point about these results, is that they didn't suddenly jump up in 2010. Instead, the rise was gradual. What that indicates is that more schools, and more teachers, began gradually to get better at, and more committed to, teaching phonics well, over a period of years.
This is what is wrong with the argument used by Bojo above, about phonics failing children. There is plenty of evidence that even very recently too many schools still aren't teaching phonics properly. (Unfortunately, the recent change in the format of SATS means that we no longer have that source of data.)