The logic of creating a system in which schools competed for pupils and in which heads had control of budgets - when combined with other pressures such as the high and rising cost of housing in the South-east - was that each school would eventually pay teachers largely according to the dictates of the market. If a head spots a potentially brilliant maths teacher working for a bank, why should she not pay a premium rate to attract him into education, and cut the number of teaching assistants or the equipment budget in order to pay his salary?
Equally, if she has a history teacher who is boring children out of their minds and (in her view) depressing exam results, why should she give him an automatic annual rise? If heads are to be held responsible for their schools' results, it is hard to deny them the right to spend money and reward teachers as they wish.
It has taken a decade-and-a-half and a Labour government for my prediction to come anywhere near fulfilment. Ministers thought they were boxing clever on performance-related pay. The strategy was to introduce it for teachers at the top of the scale, but to put so much money on the table that, provided they completed a long and fiendishly complex set of forms, more than 90 per cent would be successful. The unions were thus cornered. To oppose PRP was to oppose a big pay rise for a large section of their members as well as to deny a process whereby it would be officially confirmed that nearly all teachers were brilliant, as the unions had always insisted.
The thin end of the wedge thus firmly inserted, ministers could then insinuate PRP across the profession by concentrating the lion's share of future pay awards on those above the threshold. But that would be no good if crossing the threshold, and proceeding further above it, became just like getting another annual increment. So heads and teachers had to be persuaded that PRP entailed failure and disappointment as well as success. In the second year, therefore, ministers offered schools a sum of money that was quite inadequate to reward all teachers above the threshold who could potentially move up an increment.
Headteachers were supposed to select, to distinguish between the brilliant and the not-quite-so-brilliant. They would have none of it. A PRP scheme that could not reward almost everybody was, they argued, inherently absurd. So ministers have been compelled, once more, to make the explicit case for PRP. They have been outmanoeuvred.
But the attempt to introduce PRP is a fool's errand. It involves too many forms and a preposterous superstructure of inspection. It is a perfect example of the mistake that governments have been making for nearly 20 years. They want the public sector to operate more like the private, but they also want to retain control; they want flexibility and diversity, but they do not want to give up the old public-sector command mechanisms.
Most private sector firms do not have PRP and the few that have tried it have generally given it up. They leave it to the boss to decide who deserves a pay rise; and the boss, without the help of forms or dossiers, usually knows who's working hard, who's making himself indispensable and who's likely to be tempted elsewhere. That system is alien to the public sector, which insists on accountability at every level.
I make here no judgment about which is the better system: indeed, I dare to suggest that both have merit, depending on circumstances.
But I am fairly sure that they cannot be combined. If ministers really want market-responsive schools, and teachers who are paid according to merit and market value, they should scrap all national pay scales and set the heads free.
Peter Wilby is editor of the New Statesman