By Mike Baker
BBC News education correspondent
What makes a good school?
This question, so often asked by parents, goes to the heart of the current debate over the government's reforms for England's secondary schools.
That is because the government's proposals are driven by Tony Blair's concept of a good school.
We need to understand this to make sense of the government's plans.
In the foreword to the White Paper, the Prime Minister wrote that the "best state schools" were those with a special ethos, a clear sense of purpose, strong leadership, energetic sponsors, and motivated parents and pupils.
How do you define good?
Just hold onto that for a moment while we look at other definitions of good schools.
We could turn to the newly published 2006 edition of The Good Schools Guide. It is a consumer-driven list of 800 "good" state and private schools chosen on the basis of parents' comments.
The choices are revealing. I looked up the selection of schools from one county, chosen at random.
There are 10 Essex schools in the book. Four are fee-charging schools. Three are selective state grammar schools. That leaves just three schools that are open to all, irrespective of entrance exam or ability to pay. All three are foundation schools that have control over their admissions.
One of these is a "European" school. Its admissions arrangements are complicated, with priorities given to some local children but also to those with a special connection to the school's European ethos.
For example, 10% of places are offered to those who have a parent who has "visited a country other than the UK through their work or interests".
The plans would affect England's secondary schools
So that leaves just two "good schools" in Essex that operate admissions based purely on local catchment area.
Both are in fairly affluent areas and, according to Ofsted, both have well below the national average proportion of pupils eligible for free school meals or with special educational needs.
What does this suggest? If the book's selection does indeed reflect the public's view, it seems that parents regard a school as "good" if it is either competitive to get into or has a fairly homogenous middle-class pupil intake. In other words, the key is the school's intake or admissions.
So let's try another definition: the 200 top-scoring comprehensives at GCSE. Research this week from the Sutton Trust found that, on average, these schools had only one-third the national average of pupils from families poor enough to be eligible for free school meals.
In other words, they tended to be in more affluent areas. More significantly, they had only about half the proportion of poorer pupils that lived in their local area.
In short, many of these comprehensives were socially selective in some form or another, whether by religion or motivation.
About 70% of these schools are either religious or foundation schools that operate their own admissions, in much the same way as the government plans to allow its proposed trust schools to do.
While these schools are not selecting by outright academic ability, or by social class, they are getting the most active, committed, and skilled pupils from the most supportive and involved homes
If you look up the admissions criteria at some of these schools, you see how this selection - while not ostensibly choosing the brightest pupils - effectively prioritises the more motivated pupils from the more active, energetic and supportive homes.
For example, one Catholic school not only prioritises applicants who make the school their first preference and who are practising Catholics, but also asks them how often they and their children attend Mass, how actively involved they are in the church as, for example, readers, choristers, or members of the union of Catholic mothers.
Another non-denominational Christian school gives priority to pupils who can show active involvement, over a long period, in activities such as choirs, orchestras, sports teams, brownies and cubs.
So, while these schools are not selecting by outright academic ability, or by social class, they are getting the most active, committed, and skilled pupils from the most supportive and involved homes.
It is little wonder that these schools emerge as "good" schools on almost any criteria. I have visited many of them and they are impressive.
Now the government has looked at these schools and concluded, reasonably, it wants more of them.
But the tough question is this: can every school be like these schools without, in some subtle way, selecting its pupils?
One can only assume the government regards this element of non-academic selection as a prerequisite for success. Otherwise, why does it want trust schools to take control of their own admissions?
This is the point where the government's critics, including the Commons education select committee, warn that trust schools will lead to a more segregated system.
In its report, the select committee raised the concern that "if more schools act as their own admissions authorities they will use that power to choose pupils who are likely to perform well academically".
It cited the advice of the chief schools adjudicator (the man in charge of monitoring school admissions) who argued that popular schools tend to "drift upmarket".
Judging from the way it is dividing the Labour Party, it may regret letting the genie of selective admissions out of the bottle
Research from the London School of Economics confirms this, suggesting that schools that have control over their own admissions tend to have fewer pupils from poor homes or with special educational needs.
Just to be clear, this is not old-style 11-plus-style selection by academic ability. Nor is it as crude as selection by social class. No, this is selection by motivation and parental ability to understand and "play" the system.
It is hard to blame parents for trying to do this. But critics of the school reforms say it casts doubt on whether you can make all schools good by giving them this freedom over admissions.
They argue that for every school that is able to use its admissions to admit the more motivated pupils and parents, there will be another that loses them.
This is why the education select committee wants to give greater powers to local authorities to ensure fairness in admissions across a whole local area.
However, supporters of the government's changes point out that the current system is hardly fair either. Admissions based on catchment areas favour those who can afford to buy a house close to a good school.
Similarly, supporters of academic selection point out that the grammar schools provided the best ladder of social mobility for bright disadvantaged pupils this country ever had.
Both are right. The difficult question is this: which system produces the greatest possible number of good schools?
If a "good school" is one which takes a disproportionate number of pupils from the most supportive and committed homes, surely that will lead to other schools being deprived of such pupils?
The BBC school league tables suggest that the areas with the highest proportion of both high and low scoring schools are those where there is a very high proportion of schools running their own selective admissions policies, places like Lincolnshire and Kent.
These counties have both excellent schools and some of the lowest-achieving schools in the country.
With the latest political row over its reforms, the government has done exactly what it tried to avoid doing immediately after 1997. It has focused attention on admissions and on types of schools.
Back in 1997, it argued that the way forward was to focus on "standards not structures". Now it is, almost unwittingly, focusing on structures.
Judging from the way it is dividing the Labour Party, it may regret letting the genie of selective admissions out of the bottle.
We welcome your comments. A selection will be published later.
The BBC may edit your comments and not all emails will be published. Your comments may be published on any BBC media worldwide.