First, scale:

A district like Seattle (with which we work) has about 60,000 students. At about $6000 per student per year that is a total operating budget of about $360 million dollars a year.

Corporations typically spend on the order of 10% for training, R&D etc... Microsoft spends 16% of its funds on making itself better...

So lets take 10% as a figure... That means that if we are serious about real improvement in our school systems, then every level of the system would need 10% of its net revenue devoted not to running the system, but solely devoted to having the capacity to improve itself...

For Seattle that would mean having 36 million dollars each year devoted in a line item to the improvement and maintenance of instructional programs...

Roughly that might mean 18 million for elementary and 18 million for secondary... If you assume that elementary science is one-fifth (or even one eighth) of the elementary program... then Seattle would devote about $2 million dollars per year to having a permanent capacity for improving elementary science... for maintaining and supporting a strong elementary science program... This amount of money we should note is slightly more than what the current NSF LSC provides...So in a way what is happening is that NSF funds are providing Seattle with a temporary improvement infrastructure... The NSF funds approximate the kind of work and the kind of capacity that should be there all the time -- IF WE ARE SERIOUS (BEYOND RHETORIC) IN IMPROVING THE QUALITY OF SCIENCE EDUCATION... Hence we need to somehow conceptualize and then make real the idea of ongoing investment in internal capacity for continuous improvement... Right now improvement is a soft money and temporary exercise... And like the Golden Gate Bridge the most steady forces are those of erosion...We know they will not go away.. and so we need similar steady forces of restoration and improvement...

And what of the components of the improvement infrastructure?

Well, again, I think the LSCs provide a good first approximation... First and foremost they provide a point person... a full time leader and advocate who oversees the health of the improvement effort... Second they provide for TOSAs ... teachers on special assignment who do all the real work of inservice and support... and in turn receive wonderful leadership training...Ane there are funds for outside experts to help the district design and implement and evaluate its improvement efforts... Third the improvement infrastructure has to handle the physical and logistical needs of teachers ... the materials, the replenishment etc... Fourth the improvement infrastructure works hard to provide a supportive context for the improvement and maintenance efforts... from parent and public understanding to district administrative support and alignment... The improvement infrastructure also provides important connections to national and state resources ... And Fifth, the infrastructure has its own RandD so that it is not only maintaining good programs but it is always working on the next steps for the district in elementary science, or middle school math or whatever the domain is...

But before any of this can happen we have to conceptualize it... and argue for it... it is true that all the funds in the system are not even sufficient to run the system well...Surely additional funds are needed...But as additional funds come... we must have in mind the goal of building this improvement infrastructure at all levels... Currently we do not think this way at all...