G120. Its terms of reference asked the review to look specifically at a system of banding for NNDR, as a means of improving the certainty, stability and simplicity of revaluation for ratepayers, while maintaining an appropriate level of fairness in the relative valuations of properties. The main areas considered were:
- Would banding improve the simplicity and fairness of the rating system;
What number and width of bands would be required; What would the effect of a revaluation be on banding and rate bills. G121. The consultation responses were overwhelmingly opposed to banding, mainly on grounds of fairness. 93 out of 114 were against banding, while just 7 were in favour.
Would banding improve the simplicity and fairness of the rating system?
G122. Banding would appear to have considerable advantages in terms of certainty, stability and simplicity. A change of value within a band would not affect a ratepayer’s bill, only a movement between bands. As a result, there would be less need to undertake revaluations in order to keep the lists up to date, so there would be greater stability and certainty. With fewer revaluations and changes in bills resulting only from movements between bands, there may be less need for transitional relief and therefore greater simplicity. Only ratepayers who thought their values were in the bottom margin of a band would have an incentive to appeal, thus improving certainty for ratepayers and for central government. However, a closer analysis suggests certain reservations on all these points.
G123. Having fewer, broader bands would increase the extent to which these objectives were achieved, compared with more, narrower bands. However, the cruder the system the more difficult it would be to operate it in a way that maintained an appropriate level of fairness. There would be gainers and losers under any banding system and there would be more of these within a system of broad banding. Broader bands would also mean higher rises in bills between bands.
G124. In considering the fairness of a banding system, the limitations of the existing system also need to be borne in mind. Valuation is not a precise science. Under the present system, many - if not most - of the rateable values are arrived at by a process of negotiation between the valuation officer and the ratepayer, with valuations being settled following an appeal. The appeal is either decided by a valuation tribunal or settled by the parties before it reaches the tribunal. A system of banding would recognise explicitly that the valuations are not precise. To that extent it need be no less fair.
G125. Council tax provides a model of simplicity and stability in a banding system. But the position is different for non-domestic property. There are just 8 council tax bands, but this is unlikely to be sufficient for non-domestic property. A significant difference is that council tax is based on capital values, whereas rates are based on annual rental levels, with a very different distribution of values, ranging from £1 to hundreds of millions. There is also a fairly narrow range of tax levels for council tax, with a band H bill being three times a band A bill. It would be impossible to keep such a short range to cover the widely differing values of non-domestic property, while maintaining the same level of total yield.
G126. As a result, even a marginal increase in value that moved a non-domestic ratepayer from one band to the next could lead to a significant increase in bills, probably much more so than under council tax. It would be unfair for near neighbours or competitors in similar properties to find themselves paying markedly different rate bills as a result of banding, when they pay similar amounts under the present system. The example bands below show that a difference of one band could mean substantial increases in the bill paid.
G127. It has also been suggested that banding provided a way of simplifying the system by reducing the scope for appeal, if there was little likelihood of a successful appeal resulting in band movement. However, responses to consultation from the rating profession suggest that there may be greater incentive to appeal because of the implications of a large reduction in liability following a band reduction. There may also be a disincentive to settle an appeal.
What number and width of bands would be required?
G128. An example of possible bands is shown below. This is just one option, based on a reasonably even distribution of the number of properties and of the proportion of total rateable value between each band. There are 13 bands, few enough for the system to be simple, but sufficient to cover the range of values. It would of course be possible to set the band boundaries differently and to have a different number of bands.
G129. Properties with very large rateable values (£200,000 and above in this option) would still probably need to be valued individually. This group covers 1.5 per cent of properties and is worth 38.5 per cent of total rateable value. There is a wide spread of values for this relatively small number of properties, so it would be hard to devise meaningful bands, while maintaining an equitable contribution to a fair share of total yield.
Band
Value (£ RV)
% of all Properties
% of Total RV
Possible bill
A
Under 1,000
7.6
0.2
£208
B
1,000 to 2,499
15.8
1.2
£728
C
2,500 to 3,999
13.8
1.9
£1,352
D
4,000 to 5,999
13.3
2.9
£2,080
E
6,000 to 7,999
9.1
2.8
£2,912
F
8,000 to 10,999
8.8
3.6
£3,952
G
11,000 to 14,999
7.1
4.1
£5,408
H
15,000 to 19,999
5.6
4.2
£7,280
I
20,000 to 29,999
6.0
6.5
£10,400
J
30,000 to 49,999
5.2
8.7
£16,640
K
50,000 to 74,999
2.6
7.2
£26,000
L
75,000 to 99,999
1.4
5.3
£36,400
M
100,000 to 199,999
2.1
12.9
£62,400
G130. The level of bill shown in the table assumes that the multiplier of 41.6p for 2000/01 is applied to the value at the mid-point of the band, for the purposes of illustration only. Alternatively, the multiplier could be applied to the top or the bottom of the band. Whichever approach was taken, this example makes clear that a difference of one band makes a very substantial difference to rate bills. This reflects on the concerns about the fairness of the system to similar properties on either side of a band boundary.
G131. An alternative approach would be to define a fixed ratio of the bill for each band, as is done for council tax. But this would probably not eliminate the steep step changes between bands, given the wide range of values that are covered, if total yield is to be maintained. The only way to reduce the large steps between bands would be to have a much larger number of narrow bands. But as noted above, this would result in a banding system very little different from the current system of individual valuations.
G132. There would also be winners and losers in each band, compared with the current system. Those whose current value is towards the bottom of each band would pay more than they do at present, while those towards the top would pay less. The proportion of winners and losers would depend on exactly how billing was calculated in a banding system.
G133. In practice, a different multiplier would be needed to maintain constant yield, depending on which point of the band it was applied. The multiplier would need to be lower if it applied to the top of the band and higher if it applied to the bottom, in order to produce constant yield. The amount paid in any band would be different, but is unlikely to be significantly so, if total yield is to be maintained.
Banding and Revaluation
G134. A further issue is what would happen at revaluation under a banding system. This raises questions about whether or not the bands themselves should be amended in line with the revaluation effect. If, for the purposes of illustration, we assume that a banding system had been employed at Revaluation 2000 and those illustrated above were the result of a 25% increase across the board, reflecting the total increase in rateable value in England, then the pre-revaluation bands would have been as follows:-
Band
Value (£ RV)
A
Under 800
B
800 to 1,999
C
2,000 to 3,199
D
3,200 to 4,799
E
4,800 to 6,399
F
6,400 to 8,799
G
8,800 to 11,999
H
12,000 to 15,999
I
16,000 to 23,999
J
24,000 to 39,999
K
40,000 to 59,999
L
60,000 to 79,999
M
80,000 to 159,999
G135. Had these bands applied before the revaluation, and the bands above which are 25% higher applied afterwards, the following results would have been seen:-
- 52% of properties would have stayed in the same band;
- 9% of properties would have moved up one or more bands;
- 38% of properties would have moved down one or more bands;
- 81% of those moving would have moved by just one band
- The upper and lower bands are more stable, with 89% of band A properties and 61% of band M properties not changing bands;
- The middle bands are less stable, with 32% of band H and 35% of band G and L properties not changing bands.
- A very small number of properties would have moved a large number of bands.
G136. These figures reflect the fact that at the Revaluation 2000, there was a significant number of properties with changes in rateable value at the lower end of the spectrum and a smaller number of high value properties seeing large changes in rateable value.
G137. An alternative approach would have been to apply the same bands before and after revaluation. On this assumption the results would have been similar, with the following effects:
- 61% of properties would have stayed in the same property band;
- 27% would have moved up at least one band;
- 12% would have moved down at least one band;
- 82% of properties that moved a band only did so by one band;
- The upper and lower bands would be more stable, 80% of properties would have remained in band A and 68% in band M;
- The centre bands appear less stable with, 47% of properties remaining in band H and 41% in band L;
- A very small number of properties would have moved a large number of bands.
G138. Appendix 5 contains more detailed data on these two scenarios. Whichever approach is taken to setting bands, a sizeable number of properties would move bands during a five yearly revaluation, particularly in the middle of the range, but for the vast majority of these properties it would have been restricted to a movement of one band. However, as noted above, movements of one band can result in significant changes in the amount paid in rates.
G139. The main difference between the two approaches taken is in the numbers moving up or down the bands. If the bands were constant, 27% would have moved up and 12% would have moved down, a total of 39%. If the bands had been uprated, 9% would have moved up and 38% would have moved down, a total of 47%.
G140. It may be that, with an uprating of bands, a banding system would produce more manageable stability at revaluation. Most would face a modest increase in their bill, as they stayed in the same, uprated band. This assumes that any multiplier was reduced to reflect the increase in values and bands, but the effect was not complete because of the allowance for inflation and losses on appeal that are included in the multiplier (see below). 38% would face lower bills as they would be in a lower band, while only 9% would face the substantially higher bills resulting from an increased banding. This compares with the 40% of ratepayers in upward transition in 2000/01 under the current system. But this is just one example and further analysis would be required to be sure that this one example could be relied upon to draw conclusions about the effects of a banding system at future revaluations.
G141. The analysis above of the effect of a longer revaluation cycle on rateable values suggests that there would usually be more movement than with the five-year cycle illustrated here. This suggests that under a banding system, there would also be some more turbulence with a longer cycle.
Conclusions
G142. A banding system appears to have advantages, particularly in terms of simplicity and, subject to further analysis, perhaps also stability. But there are significant concerns about its fairness, particularly because of the difference between bills in adjacent bands. This was a particular concern of business respondents to the consultation. Depending on where the lines are drawn, the step changes in rate liability as properties move from one band to another could be significant. Ratepayers are likely to regard it as unfair that a small movement in value could affect their bills to such a significant extent. It could also have an adverse impact on investment decisions, and act as a real disincentive to the improvement or expansion of business premises.
G143. There are significant differences from the council tax model, which mean that it would be impossible to avoid such step changes between bands, without having so many bands that the system is virtually the same as one of individual valuations. For practical reasons, it would probably be necessary to leave out of banding the highest value properties, which account for a small number but a large proportion of total rateable value. It was not clear to the Review Group that any obvious advantages would accrue from the introduction of banding for non-domestic properties at this stage.
[ Annex G Section4 ] [ Contents ] [ Annex G Section6 ]
Published on 19 September 2000
Return to Local Government Finance Index
Return to Local Government and the Regions Index
Return to DTLR Home Page
DTLR Web Site Terms