There are many good reasons for setting targets. Local authorities have priorities besides education - getting rid of rubbish, making sure streets are in good repair, providing elderly and social care.
The argument goes, that without the incentive of having to justify themselves to the Government every year, providing quality early years education could slip down the list. Targets also have an important role in making it transparent to all how public services are measured.
But there are concerns about target setting. Before all else, the data on which the targets are set has to be reliable. The first foundation stage profile results published in 2003 were not considered robust enough to be National Statistics. Since then money, time and training has been pumped into moderation and each subsequent year the data has been declared "robust". It may well become possible to ensure that one child who can "sing a simple song" is not expected to memorise "Yellow Submarine" while another pupil hits the mark with "Wind the Bobbin Up".
But this will do little to allay the larger concerns about "juking the stats", as hit TV show The Wire describes teaching to the test.
It will not happen in early years, the Government had assured us, because no school or child-level targets will be set. But heads are expected to be take part in the authority's plans for meeting its targets to raise the proportion of children with good language and social development.
And reception teachers are also expected to set "process-based" targets for the children which fit into the play-based curriculum.
Who would argue with improving children's language and social skills, especially when schools have the information and means to do it? But these targets focus only on the benefits - they do not tell us the potential cost.