It seems almost universal for people to get a probable birth year from census data by subtracting the age from the census year. Certainly some websites and some census CDs provide this as "helpful" information. For me it isn't helpful and personally I always subtract one more year from the result.
All the 19th century censuses except 1841 were taken the first weekend in April, so people were three times more likely to have been born a year earlier. E.g. if a person was 24 in 1851, the accepted guess by subtracting one from the other seems to be 1827 for the birth year. But in point of fact, making of course the big assumption that the age is correct, this person was probably born between April 1826 and the end of March 1827. Born any later and (s)he would have been only 23 at the 1851 census. So for 1851 I subtract the ages from 1850, and similarly for the other censuses. 24 from 1850 gives 1826 which is the true "most probable birth year".
If you don't do this you may be looking in the wrong year when you go looking for details in the parish registers. Most baptisms took place before the child was three months old, and usually less than that. For someone who was "24" at the 1851 census, even the baptism may have taken place as early as April or May 1826, so I'd always start looking in 1826 and only go on to 1827 if I couldn't find what I was looking for.
In 1841 the census was taken in June, and ages above 15 are usually in a state of confusion reflecting the enumerators' understanding, or more usually lack of understanding, of the guidelines. Ages above 15 were supposed to be rounded down to the nearest multiple of 5, so that "20" stands for anything between 20 and 24 and so on. But many enumerators gave up, ignored the instructions and gave the exact age - and I always offer up silent thanks when they did!
But for ages under 15, or where the enumerator gave an age that was not a multiple of five, the birth still has seven chances out of twelve (and therefore still slightly more than 50%) of having taken place in the year before the one you get by merely subtracting the age from the census year.
Source: Tony Woodward (GOONS)