Excitement has broken out at ConservativeHome over Conservative Councils charging lower Council tax. This is measured by Band D, the “average”. Labour’s response is that Band D isn’t a fair comparison, and what should be used is the figure for how much the average house pays. The Conservatives say that using that comparison is “thoroughly dishonest… because you have huge fluctuations in property prices between different authorities“.
At risk of getting my graphs out again, I think both parties are behaving unhelpfully here. Band D certainly isn’t a real average in any meaningful sense – the average house in England is actually something like B and three quarters. So to an extent Labour have a point. Imagine for the sake of argument two Councils with only one house each in them. Poshville has a mansion house in the grounds of the Town Hall, which is in Band H, and charges the resident £1500 a year. Slumton has a council bedsit in the basement of the Town Hall, in Band A, and charges the resident £600 a year. Now according to the Band D principle, Poshville’s Council Tax is £750 a year, and Slumton’s is £900 a year. Would it really be fair to say that Poshville has set a lower Council tax than Slumton? All else being equal, Slumton raised less tax, and has less to spend on local services.
This is not an academic example – compare a Band F property in the Midlands with a Band G property in London, and consider what this might mean for the proportion of properties in each band in that area, whether the residents of the London studio flat would recognise a description of their local Council tax as “lower” if it was the same as that paid by the residents of the 3-bed place in the Midlands, and the total money therefore likely to be raised by a Band D rate of, say, £1000, in each of those Council areas.
Average actually charged per dwelling, however, pushes up the apparent charge for areas which have disproportionately large houses, it shouldn’t push it up for those which merely have disproportionately expensive houses, since that will “come out in the wash” by allowing a Council to set a lower Band D, and allowing the more expensive housing to catch the income back up. So, if Poshville’s mansion house is in fact rented out to five sharers, but the studio flat in Slumton is merely occupied by a young couple, we would want to look at the Council Tax paid per head. At this point we discover that each council is in fact charging the same – £300 a year, despite the huge variation in the “headline rate” and the substantial variation in the “per dwelling rate”. Score one against the Labour case as well, then.
This thought may explain a further quirk shown up by the Conservative figures, the very high figure for Band D among Lib Dem Councils could be down to the fact that Lib Dems are doing well in areas where there is high population density, but multi-occupancy housing rather than many small flats – broadly areas with a lot of students or ethnic minorities. I’m not sure, but that feels intuitively plausible.
It should also be said that £25 a month is probably a greater burden to someone living in a Council bedsit than it is to someone renting a room in a mansion, so Poshville could use the measure I ended up with in my previous post – tax per head as a percentage of median resident wage, and come out with a lower number. That would certainly be more legitimate than comparing Band D, a largely mythical average.
The more important point though, which is picked up by some of the commenters on the original article, is that Council Tax charged isn’t a terribly useful measure of anything, particularly when comparing one authority with a completely different one. The costs of delivering services differ substantially – for obvious reasons more will be spent on caring for elderly people in Eastbourne than in Cambridge, and more will be spent on tourist centres in Cambridge than in Slough. In any case, at a time when spending pressures are high but Council tax is capped, most of the locally determined difference in tax levels is likely to reflect past decisions, and therefore potential the priorities of a different party.
All else being equal, there may be a fair comparison between Councils around rate of increase, but even that will often be driven more by the Government’s funding decisions and the Council’s need to “make up the difference” than by the Council’s own choices around spending. If pushed to come up with a measure which allows comparison my instinct would be to go for aggregate spending as a percentage of resident Gross Value Added, but I can’t imagine that ever making it onto an election leaflet! (and I’m not even sure it’s collected at a more local than regional level). It could also make some councils look profligate when in fact they just have large populations who require services but don’t generate much GDP – pensioners, schoolchildren, and so on.
I think that made sense. I attribute any statistical errors I’ve made to April Foolery, on the “I was just testing” principle.