By Mike Baker BBC News education correspondent

The media over-reacts, politicians jump on bandwagons, parents get worried, and teachers become defensive.

This week, the first of the general election campaign, brought a prime example of this.

The Education Select Committee, in its last throw of this Parliament, published a report which said the figure of one in five pupils failing to reach the expected levels in reading in England was "unacceptably high".

The government took a lot of flak. Ministers clearly felt this was harsh, pointing out that since 1997 the numbers of 11-year-olds reaching the required levels in reading had risen from 67% to 83%.

It could also be argued that there was no benchmark for the numbers of children who should be at the "expected" level.

Premier League readers

Before 1997 no government had ever defined this level of reading skills. Nor were there any targets for what percentage should reach that level.

As for the question of what "expected levels" mean, no-one has ever satisfactorily answered that.

Is it the level of the average child? If so, getting 83% to that level would represent over-achievement.

Or is it the minimum level a child should reach by that age? In which case, 83% may not look quite so good, although if children with special needs (up to 20%, depending on the definition) are discounted it might be regarded as not so bad.

Opinion is divided on the best methods

Definitions aside, however, England's record on reading is actually pretty good when compared with other countries.

But while average reading standards are not as bad as the headlines may have suggested, an important question remains: why the big variation in achievement between schools and pupils?

For independent confirmation of England's good record on reading, you need only turn to the Pirls (Progress in International Reading Literacy Study) report of 2003.

This compared the reading skills of 140,000 10-year-olds in 35 countries. England ranked third, beaten only by Sweden and the Netherlands.

England not only beat similar European countries such as France, Germany and Italy but was also well ahead of the other English-speaking countries in the survey: the United States, New Zealand and Scotland.

Heated debate

However, England also had the widest span of reading abilities. The most able pupils led the world in reading skills. They are premier league readers.

But the least able readers were in the relegation zone: below their equivalents in Latvia, Lithuania, Hungary, the Czech republic, Russia, and Hong Kong.

So what is going wrong with our least able readers? Is it something to do with the way they are taught?

This is where the debate tends to get heated. Reading experts have long waged war over the relative merits of phonics, "real books" and "look-say".

The advocates of phonics (a method of decoding words by learning the sounds of groups of letters) have won most of the recent battles.

From a position of isolation 15 years ago, phonics is now back at the heart of the National Literacy Strategy, the approach which underpins the teaching of reading in English schools.

Any review should proceed very cautiously before rejecting the current mixed economy approach which, for most pupils, is getting pretty good results.

But for some that is not the end of the battle. The supporters of a particular type of phonics - synthetic phonics - believe it is the Holy Grail of literacy.

They have demonstrated considerable success in an experiment which has been running in primary schools in Clackmannanshire.

Moreover, they have been able to show particularly good results with boys and pupils from disadvantaged homes.

As the select committee concluded, synthetic phonics is certainly worthy of closer attention.

But the problem for policy-makers is that synthetic phonics is an all-or-nothing approach. It does not sit alongside other methods. It means putting all your eggs in one basket.

The essence of synthetic phonics is that, before children are even introduced to books, they go on a crash course to learn the 44 basic sounds of letters and groups of letters.

The idea is that this equips them to decode a very large number of regular sounding words.

They are then introduced to books which contain only these words and - hey, presto! - they find they can read. This can be marvellous for their confidence. I have seen it in action and it is impressive.

But champions of synthetic phonics say it is the essence of this approach that it is not mixed with other methods: it is synthetic phonics "first, fast and only".

Computer games

Still, you might say, if it has worked in the pilot projects what is the problem? Well the trouble is that there is a tendency in any pioneering project for the initial focus, enthusiasm and energy to boost results.

But it is not necessarily the same when it is rolled out over a longer period, a wider range of schools, and when the excitement and novelty have worn off.

Now, of course, it could be that synthetic phonics holds the answer to the under-performance of England's poorest readers. But to introduce it wholesale would be to take a big risk with the others who already out-perform most of the rest of the world.

And it could be that there are other explanations for England's under-performing tail of poor readers.

Again the Pirls study may offer some clues. Unsurprisingly there was a strong overlap between high reading ability and enjoyment of reading for fun outside of school.

Yet in England, 27% of 10-year-olds said they "never or almost never" read for fun outside of school. This was higher than the international average of 18%.

TV ban?

There was also a strong correlation between watching a lot of television or playing computer games and lower than average reading ability.

In England 40% of 10-year-olds said they played computer games every day, well above the international average of 26%.

Similarly, 20% of English children said they watched television or videos for five hours or more a day, compared to the international average of 12%.

It seems very likely that this 20% or so who watch a lot of television, play a lot of computer games and never read for fun represent the bulk of the "unacceptably high" figure of 20% who fail to reach the expected standards of reading in school.

This surely suggests that the key to this under-performance is not the result of England's teaching methods (which are producing results to match the best in the world) but of what is going on in the homes of a substantial minority of children.

On this basis, the most effective policy would be to ban TVs and computer game consoles in the homes of children who fail to spend at least an hour a day reading for pleasure.

Somehow I don't think any political party or government could do that.

But it does suggest that any review of reading methods should proceed very cautiously before rejecting the current mixed economy approach which, for most pupils, is getting pretty good results.

Synthetic phonics may be the answer in certain circumstances, and especially for children with little other exposure to reading, but a longer and wider pilot scheme would surely be a sensible precaution before getting every child to do synthetic phonics "first, fast and only".