BASE HEADER

I wish to voice and qualify my objections to the proposed housing development planned north of Milverton. I understand that new housing is required and that development must therefore happen somewhere, however there are clear factors which indicate that this is not appropriate land.

In the first case, this is Green Belt land. Obviously this does not entirely prevent development should (as the council has stated and acknowledged) there be 'insufficient suitable and available sites outside the greenbelt'. This prerequisite is however not met. Land east of the A452 and south of Heathcote is entirely suited to housing development and since this is the case, I would suggest that the council should stand by its own guideline.

Secondly there is the purpose of Green Belt land. The proposed area and the surrounding area are a local amenity. There is little space publicly accessible and removing this land would be highly detrimental. People run, walk dogs, children play etc. in this area.

Thirdly I am highly concerned about the council's prediction for housing requirements. I work as an applied statistical problem solving expert and mentor and I have used my experience to briefly look at what meaning can be extracted from the population-by-year raw data for Leamington I located on the council website. Since there is (visually) a population rate change in 2000, I have taken the data from 2001 to 2010 to see whether it forms a regression model from which a meaningful population (and therefore, housing) prediction can be drawn. Initially I applied a linear fit as shown below and this gave a p-value of zero for a 95% confidence, unsurprisingly showing strong statistical confidence that population increases with year. However, I then applied a quadratic (curved) fit and this also gave a p-value of zero for a 95% confidence for curvature, showing equally strong statistical confidence that the degree of population increase decays with year. Please see the graphs below.

Given that curvature has the high degree of significance above, it is incorrect to perform a linear extrapolation for future values. Doing so will be inaccurate and error will increase with time. I am highly concerned that in your leaflet, you state the need for a fixed number of extra houses per year for the next several years. This is a linear extrapolation which as I have explained above, is a statistically incorrect use of the data. The consequence of using the wrong model in this case will be the provision of houses for a population far greater than that which will actually occur. It is incidentally, entirely erroneous to use a linear relationship as a 'worst-case' scenario. If necessary a probabilistic (curved) worst-case can easily be calculated.

Finally, whilst I am not personally affected, I also note that the proposed development would be extremely detrimental to Old Milverton, something which I am sure the council will agree, should be avoided if at all possible.

I would like to reiterate that I am not opposed to necessary house building, but I do expect that Green Belt is avoided if at all possible (which it is in this case) and that housing needs are predicted accurately using the best statistical methods.

I would appreciate your views on the above points.