I set out previously some work I’ve been doing on the quality (defendability) of planning decisions, and published it as a table. I then turned it into a picture broken out by region. In showing it to a few people it’s clear that there is something interesting going on here, and not necessarily what you might think. And, in a final flourish, I mash the data with another public dataset and ask “does having a plan mean you make better decisions ?”
This post takes things a little further and turns this mash-up of public data into pictures using my favourite plotting tool ggplot2.
Comparing refusals and appeal performance
Grab this image and zoom into it
To explain what you’re looking at. The plot has been done at a high resolution that you’ll have to fly around and zoom in on. Otherwise all the council names would blur into each other. Each region appears in its own panel, with its own averages plotted in red. The x-axis represents success at appeal, with low numbers = low success. The y-axis represents the proportion of refusals, with high numbers representing many refusals. The dots also have a size (based on the ratio of appeals to refusals) and a colour (plan or no plan – more on this later).
You can immediately see big differences between the regions. London refuses about a third of all applications, the North East only 8%. And even within a region you can see big differences – the City of London just does a very different sort of work to the other London LPAs. There is also a big range between how often councils find themselves at appeal. Least often is [cough] Crawley at 8%, all the way up to poor old Broads Authority who issue a refusal and can safely put money on an appeal response.
Remember that this plot is treating planning applications as if they were tins of beans. We cannot differentiate between simple householders and major schemes. Losing an appeal on a big site in the green belt is quite different to a dormer.
We are programmed to look at plots like this and mentally divide them into quadrants, with some being “good” and others “naughty”. This is tricky. I think, putting my head slightly above the parapet, I would suggest that the lower right quadrant is “good”. These are positive LPAs who are often vindicated at appeal. In this same quadrant you might also find “weak” authorities who want an easy life and so permit everything. Probably upper left is “naughty”, because these are high refusers who don’t then back these decisions at appeal. See what you think. I’m struck by the big red splots, myself.
The biggy. Does having a plan influence decisions (and so appeals) ?
To answer this question we take the spreadsheet as set out in the previous post, and simply add a “yes” or “no” to reflect whether the LPA had a plan. This is just folding in another source of public data.
Recall I used a snapshot of the average of the data available across April 2011 to March 2013. So, given that there are Plans and Plans, I’ve judged a council to have a plan if they have a post 2004 Plan adopted before Jan 2012. I’m using a two year span for appeals, and I want to divide the councils who had a plan for the majority of the period. This divides reasonably neatly into one third with a plan against two thirds without.
Note also there are (no doubt) confounding factors in play. On the one hand the early adopters are (possibly) quite energetic planning authorities. On the other hand PINs will tell you that the early life of a plan is full of appeals as developers use the process to find out what the plan really means and how robust it is.
Then we create a two-way panel, dividing the data between region and additionally whether the council had a plan.
There are some interesting differences. In London, for example, the LPAs with plans are significantly more positive than those that don’t. And they do better at appeal. For London authorities the answer is clear – having a plan is good for you. But in the North West the opposite is true. Those with plans refuse more applications and then do worse at appeal.
So, let’s find out the “biggie”. Is there a difference between the planned and unplanned councils ?
# slightly fewer appeals against refusals (this is good !)
plan = no |appeal rate = 23.05011%
plan = yes |appeal rate = 22.84076%
# slightly more appeals won (this is good !)
plan = no | appeals won = 65.00039%
plan = yes| appeals won = 65.84087%
# but slightly higher rate of refusal (this is bad)
plan = no | appeals won = 16.54231%
plan = yes| appeals won = 17.41984%
So, in this formulation of the question the answer is (only very slightly) that plans make life better. Perhaps if we weighted the question to recognise that LPAs are not all the same size we’d get a slightly more positive result. As always, your thoughts are welcome.