This is one of two pieces on setting fees – see also the companion post on practicals. It’s written at the end of June 2011, when there hasn’t even been an announcement on localisation of fees, let alone any details. It’s based on some thoughts I’ve had and ideas I’ve stolen from front-runner work I’ve been doing with some volunteer authorities (bless you).
The post is written to help fee-setting captains process data and interview planners to inform the fee-setting models in sheets 5.1 and 5.2. BEFORE YOU START, SAVE A COPY OF THE SPREADSHEET THAT IS DIFFERENT TO THE MASTERSHEET THAT YOU USED TO CREATE CHILD TIMESHEETS.
Completing the ‘Quantities’ Column
We watch the adverts that tell us that performance histories are no guarantee of future investment yields, but it’s not a bad place to start. Here is the commentary that goes alongside the worked example:
- Begin by getting a report of last year’s applications. All valid apps, not just determined. Include anything that might help – ‘Q’ codes, fee, plot size, description, your local application type (FUL / outline / reserved) and anything else that you can use to speed up the job.
- Insert 5 columns, somewhere near the left-hand-side. Use freeze panes to keep your headings visible. I used:
- Eeek: a code I put a “1” in if I wanted a second opinion or wasn’t sure of myself given the description. Possibly I’d need to look it up or call it back. [I also used this column to put a “m” in if I thought the fee was mixed. This isn’t relevant for counting applications, but was something I wanted to think about anyway. For counting purposes, we’re only interested in the dominant part of the development.]
- Type: for standard applications this is blank. For everything else I use a selection from “outline”, “reserved”, “retro” – there may be others useful to you. Maybe you record ROMPs as applications ?
- Unit count: some applications are sized, and you should pick up how many units are being developed. Sometimes this is really easy “Convert to 6 one-bed flats”, but unless you are really lucky, this will involve you making guesses based on the description. Particularly floorspace.
- Application count: sometimes the application seeks to discharge a number of conditions or reserved matters. The fee is calculated by assessing the full fee and then discounting – use application count to record how many conditions or matters are being applied for.
- Process the data. In my world, I would start by getting my two best validators together, and give them the PAS guide to fee categories. They then separately blitz through the list of applications, filling out the ‘P’ code and any other column that’s relevant. [If this seems ‘extra’, don’t forget that they’ll also be learning how to apply the new codes from now on]As a guide, I reckon I worked at a rate of about 900 applications a day. Somegardener was faster, but then she has a core of validation skills whilst I am strictly a dilettante. You then compare and contrast between the outputs from your validators, using any differences of opinion as a way of learning the new ropes. Plus the team learn another perspective on writing development descriptions.
- Understand your “free go”s. So, at this stage you’ve got your history with new columns. Almost there. In your list of applications you will have some (maybe 20% ?) free goes. If, in future, “free go” becomes “cheap go” do you expect this number to reduce ? The easiest way of having a count is to sort your data by fee. All those zero fees ? Delete them.
- Pivot this stuff. Get a friend or look here for an overview on this most useful of tools. We’re in the realms of “easier to show than to write down”, but here goes. You need 3 x pivots.
- ROW- ‘P code’ + COLUMN ‘Type’ + DATA ‘Count of P codes’ (copy and paste this to a separate sheet)
- ROW – ‘P code’ + COLUMN blank + DATA ‘Average of unit count’ (copy and paste the output alongside your first one
- ROW – ‘P code’ + COLUMN blank + DATA ‘Count of fee’ (copy and paste the data alongside)
Here is a screenshot of my sample data. You can see that I’ve added another column – called ‘c/f’ and painted yellow. This is the data that I will carry forward into the ‘Quantity’ column in the ‘Fixed fees’ and ‘sliding scale fees’ sheets. Note that this list does not include the gaps that denote no applications, so you won’t be able to copy & paste the whole lot in one go.
A couple of points to note before we move on. Look at line for code P0100. I had 146 applications for creating (or converting to) between 1 and 9 residential units. Of these applications 114 had a fee – so 32 were ‘free go’. My ‘c/f’ figure of 130 represents my judgement of how many free goes will remain when they become ‘cheap goes’.
The average size for a development in this banding was just over 2 units – this reflects the many applications that are for single units. This is important for pricing the average applications, as your mix of applications is not normally distributed around the mid-point.
The resources column for sliding scale fees
The spreadsheet ships with some maths in it. Put in some costs into the workbook – just enough to create a sensible hourly rate. Look at sheet 5.2 and delete the formulae in cells F5 to F8. Now put 50 hours in cell F5. Because the bands relate to their neighbours, you’ll have a cost per unit that is negative in the band below. It’s plain confusing to have a spreadsheet echo negative numbers back to you. That’s why it ships with an assumption that planning costs reduce as you move through the bands – and the figure used is 25%.
This 25% ‘economy of scale’ may not be aggressive enough – especially for the area-based applications in category 2. This may also be true for categories 3 and 4 but don’t have any results from Counties yet. Even if you use the maths as a guide, I reckon you need to reduce by nearly 50% to generate useful indicators. I’ll send a revised version to the registered user to demonstrate what I mean.
Note also that the ‘resources’ you apply to banded applications represent the amount of work you do at the top of the band. **NOT** the mid-point, despite this being our message at the fee events. It is just too difficult to explain and use, and the top is simpler.
Tips on completing the resources column
You will already know that you cannot populate the ‘resources’ column by yourself. You’ll already have booked some time with a super planner – someone who is prepared to engage pragmatically and make some rules of thumb appraisals of the various application types. Having done a couple of these things now, I’d offer you the following tips:
- It’s not just hours: we talk about ‘how many hours’ per application, but don’t forget the hourly rate covers more than just staff time. It is an aggregate cost, including the additional advice you buy in to cover the more complex applications, adverts, postage etc. You should reflect the balance of these costs between the simple and complex in addition to staff time in order to reflect genuine cost.
- Use your own data: You’ve just processed last year’s data to give you some quantities. When you’re discussing a particular code, flick back to it and use a filter to find applications that fall into that category. You are not “guessing” – you’ve already done applications of this sort and the descriptions / addresses will jog peoples memories. Your teamleaders in particular will know how much work goes into these things – that is after all what they do for you.
- Use several people from different roles: Even the “super planner” will not know everything about everything. I suspect I’ve been given the occasional piece of duff advice from people who no longer get involved in the simpler stuff. The people who can advise you about major majors are different to those people who deal with certificates of lawfulness or time extensions. Triangulate. Why not use your timesheet data to double-check your assumptions about validation time, or time taken to generate decision and notification notices ?
- Get going, and refer back and sideways: Most people will hesitate on offering resourcing opinions – we’re just not used to thinking that way. Allow people to think out-loud, and be prepared for the first couple of categories to take a while. But once you’ve gone a while, people find it easier to make comparative judgements – if type ‘1’ is 8 hours, then this is slightly less so let’s call it a ‘7’.
- Work through it in order: The workbook splits the inputs depending on whether the price is fixed or variable. I found it simpler to work through each category, so you’ll need to flit between fixed and variable using the ‘summary’ sheet to ensure you complete all “residential” before moving onto “non-residential”. Note that the summary sheet is a non-editable overview of your schedule – don’t be tempted to try and change it directly.
- Don’t get bogged down: I can’t pretend this stuff is easy. Chances are you will get stuck or end up in an argument of some sort – either by trying to push things along or because strong-minded people who care can be slightly awkward at times. Put it aside. Make tea. Carry on. Revisit later.
- Have a few prompt questions clear in your own mind: People will not just shout numbers out – you need to help them feel comfortable with the broader process. Are they helping you understand the whole part or just their area ? For a standard application, what sort of things do they do ? How often do applications go wrong ? Are there fairly well-understood processes to take things to committee ? if they had to choose between 10 applications of type ‘1’ and 10 of type ‘2’ which pile would they choose to give themselves an easier life ? Perhaps if someone newer was doing the application, would they take a bit longer ?