G24. The review considered whether there was justification for a move away from the current five yearly cycle for revaluations. The main options were:
- annual rolling revaluations;
full annual revaluations; three-yearly revaluations; no change, retaining the current five-year cycle; less frequent revaluations, canvassing views on a ten-yearly cycle. Some consultation respondents also suggested a seven-year cycle. G25. It became clear that there are tensions between the objectives of stability and certainty on the one hand, and fairness on the other. The revaluation process itself inevitably causes instability in rate bills. Its purpose is to redistribute the rate burden in line with the property market, to reflect changes since the last revaluation, and those changes will be reflected in individual rate bills. There will also be uncertainty about the effect of any revaluation on bills generally and individually. So the more frequently revaluations are held, the more often that instability and uncertainty occurs, while less frequent revaluations mean a greater period of stability and certainty between revaluations.
G26. On the other hand, revaluations enhance fairness by ensuring that rateable values and the distribution of rate bills based on them remain in line with trends in the property market. More frequent revaluations are fairer, because they keep rateable values closer to actual market values. Similarly, the longer the period between revaluation, the less fair become the rateable values on which rate bills are based. Under the current system, the 1995 revaluation gave rateable values as at 1 April 1993, which were the basis of bills from 1995/96 until 1999/2000.
G27. Views on whether a shorter or longer revaluation cycle would be better than the current five-year cycle depends on the relative weight given to stability and certainty, as opposed to fairness.
G28. The length of the cycle does not affect the objective of simplicity in the system. However, a move to rolling revaluations was also considered, which would be a more complex and less transparent method.
Discussion
G29. It is not the absolute level of rents in any one year that drives changes in rate bills at a revaluation. If all rateable values rise or fall by broadly the same level, however great, then the multiplier falls or rises accordingly, to keep total yield constant, and rate bills will not change significantly for most ratepayers. However, that does not happen, because the property market across England is not homogenous. Values for different localities and types of property move at different rates and even at times in different directions. So rateable values change by different amounts for different properties, with significant variances between sectors and between - and within - regions. It is these variances which lead to changes in most individual rate bills, within a constant total tax yield.
G30. The tables in Appendix 3 give an analysis of annual changes in rental value and the resulting effect of different revaluation cycles. The first table is based on data produced by Investment Property Databank (IPD). It provides an index of annual rental values for each year from 1980 to 1999, for the UK as a whole, broken down by retail, office and industrial property of investment quality (data for England only is not readily available). This enables a comparison over different time intervals - 1, 3, 5, 8 and 10 years. For each of these the table compares changes in rental values in any year compared with a given number of years previously. This shows the variances that would occur between the valuations of properties in different sectors, though it does not show variances between regions.
G31. The second table is based on data from the VOA’s property market report, from 1988 to 1999. It shows rental levels for shops, offices and industrial property in four English locations - Manchester, Birmingham, Bristol and Camden. Again, it compares changes in rental levels over 1, 3, 5, 8 and 10 year periods. This snapshot of four locations illustrates the variations between them that would occur at any given revaluation on any given cycle.
G32. Both the IPD and VOA data concentrate on good quality property, where the market is most active. However, NNDR is payable on all properties, including those that the property market considers to be of secondary quality. Market trends for these secondary properties, which account for a large proportion of the total, are likely to be different from those shown in the IPD and VOA data, further adding to the variances in revaluation effects for any given revaluation cycle.
G33. The case for shorter revaluation cycles is based on the assumption that there will be less turbulence at revaluation because there is less fluctuation in the property market over a shorter period. The tables in appendix 3 confirm that changes in value are less over a shorter period and more over a longer period. However, they also show that there can still be significant variances in value changes and resulting turbulence in rate bills on a three-year or one-year revaluation cycle.
G34. More frequent revaluations will of course be more resource-intensive than less frequent revaluations. Shorter cycles may require more effort to complete the revaluation process in a shorter timescale, for example a two-year revaluation period would not be practical for a one-year cycle and may be too long for a three-year cycle. Even if the same resources were employed at each revaluation, costs would be greater over time simply as a result of there being more revaluations. Conversely, less frequent revaluations would lead to lower costs.
G35. Revaluation cycles may also have an effect on the level of appeals. This in turn affects both loss of yield as a result of appeals, and the cost of running the appeals system. The presumption is that more frequent revaluations lead to less turbulence in rateable values, and so to fewer appeals, while those that go forward result in a smaller loss of yield. Even if this is the case, this may be offset by the increased frequency of revaluations meaning that the appeals cycle also occurs more often. So, over a ten-year period, three three-year revaluations may lead to fewer appeals each time, but more in total, than two five-year revaluations or one ten-year revaluation.
G36. In any case, there is no firm evidence on which to base any assumptions about the effect of revaluation cycles on appeal rates. There is currently an appeals culture, so many ratepayers, or their agents, appeal whatever their valuation. On the 1995 lists, over 1 million appeals have been made against 1.6 million assessments in England. Some of these are multiple appeals and for properties with rateable value over £15,000 there have been more appeals than there are hereditaments. Unless this appeal culture can be changed, different revaluation cycles are unlikely to affect appeal rates. The changes to appeal procedures introduced on 1 April 2000 aim to address this, by establishing a published programme for dealing with appeals. A better exchange of information between ratepayers and the VOA in the course of revaluation, whenever it takes place, would also help and this is considered elsewhere in this annex.
Annual revaluation
G37. Three types of annual revaluation were considered by the Review:
- a full annual revaluation of all properties;
- a "rolling" system whereby one-fifth of all properties would be revalued each year and used as an index by which to update all other values annually;
- a staggered revaluation where properties of different types would be valued each year.
G38. Only 5 consultation respondents out of 114 supported annual or rolling revaluations. The RICS also showed some interest in annual revaluations, but in view of the practical constraints favoured three-year cycles.
G39. As noted above, annual revaluations would deliver purity in the system and the fairest revaluation cycle, as each year every property would be given a new rateable value which reflected an up to date current market rental level. With a five-year cycle, properties have to wait five years for any changes in their actual value to be reflected in their rate bills. While this benefits those whose values are rising over this period, it is unfair on those whose values are falling and have to wait to see the benefit.
G40. However, most rent reviews are on a five-year cycle and are not reviewed annually, even though annual rent reviews would better reflect actual market values and so be fairer. Most commercial tenants are prepared to forego the purity of an annual rent review in exchange for a period of stability and certainty in their rents.
G41. Support for annual revaluations, whether done on a full or rolling method, is also based on the presumption that they would smooth out changes in rateable values. The data in appendix 3 suggests that there would be less change than with longer revaluation cycles. However, changes in value can still be significant in some years. More significantly, there can also be substantial variations between regions and sectors in some years, which will lead to large increases and decreases in rate bills.
G42. The IPD data show that on an annual cycle, changes in value are relatively small in most years, usually less than 10%, though in some years this is a decrease rather than an increase. However, there are increases of up to 25% (in 1987 and 1988) and decreases of up to 20% (in 1992). Variances between sectors are between 3% and 7% in most years, but reach 13% in 1987 and 1991 and 17% in 1992.
G43. The VOA data show that there is more turbulence when comparing regional variations on an annual cycle. While in many years there is no or very little change, in others it can be substantial. For example, the value of shops increased by 43% in Birmingham in spring 1997 and by 35% in Manchester in spring 1998. The variance between regions and sectors is never less than 16% and is 35% or more in 12 of the 22 time periods, reaching 81% in autumn 1990. While these variances are less than for longer cycles, this data suggests that a move to annual revaluations will not guarantee stability in rate bills.
G44. Annual or rolling revaluations would expose ratepayers more fully to the ups and downs of the property cycle, with each annual shift feeding directly into rateable values, giving ratepayers much less predictability and stability than at present. While the effect across England as a whole would be smoothed by annual changes in the multiplier, there would be no certainty from year to year as to the extent or even direction of changes in individual rate bills, which could go down one year and up the next. Transitional relief would not be available to cushion the effects of any substantial increases or decreases in bills, as it would not be practical to devise a scheme that applied within a single year.
G45. Over a period of five years, rateable values and the multiplier should be broadly the same whether there had been annual revaluations or a single revaluation in the fifth year. But the steps in between would be very different. Annual revaluations would leave ratepayers open to unpredictable and perhaps volatile changes in RV, the multiplier and their bills in every year. The current five year cycle would deliver stable RVs and real terms multiplier throughout the five years, with any large changes phased in by TR in fixed steps which would all be known in advance from the start of the scheme.
G46. The current revaluation process takes more than one year to complete. Even with the changes to the timetable outlined above, a full revaluation would be implemented two years after the valuation date. This may be further reduced if more resources were employed, but there are limits to what is possible within a year. So a full annual revaluation of all properties would probably need a more mathematical process than the current system, which has developed to reflect quite small variances between the values of similar properties. A move to annual revaluations, with less time for such finessing, might require the acceptance of a broader brush approach.
G47. Ratepayers may wish to appeal against their new rateable values every year under an annual system, which would also have considerable resource implications in terms of handling those appeals. At present, it often takes more than a year to deal with an appeal case, though the reforms to the appeal system introduced on 1 April 2000 should help with the management of the current level of appeals. On the assumption that annual revaluations deliver lesser fluctuations in bills and in values, there may be fewer appeals in total, though there is no evidence to support this view.
Rolling revaluations
G48. An alternative suggestion is for a rolling revaluation. As with a full annual revaluation, rateable values would still be revalued every year for all properties. However, only one fifth of all properties would be subject to a full revaluation each year, with the rateable values of the remaining four-fifths adjusted by an indexation factor based on similar properties which had been valued that year. Each year a different group would be valued, so that all would be fully revalued over a five-year period.
G49. It is suggested that this would allow valuation resources to be applied more evenly. Similarly, only those properties that received a full valuation would be allowed to appeal against it. So, as at present, they could appeal once every five years, smoothing out the current workload, though not necessarily reducing it.
G50. However, the process of identifying a representative sample of properties of each type in each location throughout England, updating this each year and then calculating values for the 80% of properties not fully valued each year, may not be significantly less resource intensive than a full annual revaluation. There would still need to be an annual analysis of market rental levels in each sector and location, to value a representative 20% of properties under a rolling revaluation, as there would for all properties under a full annual revaluation.
G51. A rolling revaluation would have a similar effect on rate bills as a full annual revaluation. Assuming that a representative sample of properties could be found in each locality and for each type of property, the same variances in values would occur across England. So it would have the same advantages and disadvantages in terms of stability and certainty.
G52. However, a rolling revaluation would not have the advantages of a full annual revaluation in terms of fairness. For 4 out of every 5 years, values of any one property would be based on the values of other properties. There would be no right of appeal against those values. While they may be based on similar properties in terms of type and location, it is more arbitrary than the current system of individual valuations, which reflect all the individual characteristics of any property. Any aspects of the valuation not fully reflected by this system would be corrected every five years when a full revaluation is carried out for that property. So, as at present, a valuation reflecting the actual market value of an individual property would be given every five years.
G53. Rolling revaluations would make it harder for ratepayers to understand their bills, conflicting with another objective of the review - simplicity. Both the current five-year cycle and a full annual revaluation would give every property a value based on actual market rental levels at the time of revaluation. Rolling revaluations would break that clear and simple link. Instead, in four out of five years, ratepayers would be expected to accept a value based on properties other than their own.
Staggered revaluation
G54. The House of Commons Treasury Sub Committee, in its report on the VOA of 28 October 1999, recommended "that consideration be given to staggering the revaluation of properties according to their classification (e.g. shops, factories) in order to smooth the workload of the VOA". In its response of 26 January 1999, the Government undertook to consider this recommendation as part of this review.
G55. This is, in effect, a kind of rolling revaluation, but based simply on class of property, rather than finding a representative sample of all property types to value in each year. It shares many of the advantages and disadvantages of a rolling revaluation. It would be a more complex system than a one-off revaluation, which would undermine simplicity and transparency for ratepayers. It would also undermine stability, as it would require annual changes to the multiplier to maintain total yield, which would directly affect the bills of those classes not being revalued in a given year. This could lead to frequent and unpredictable ups and downs in bills for all ratepayers and would be unfair on those not being revalued.
Three-yearly revaluations
G56. A three-year revaluation cycle is supported by some in the rating profession and was recommended by the RICS's Bayliss Report in 1996. 29 consultation respondents out of 114 favoured a three year cycle, including 10 ratepayers and 8 rating professionals. Many of these linked a shorter revaluation cycle with reducing the need for transitional relief, though others wanted to keep TR with a three-year cycle.
G57. A three-year revaluation cycle would share some of the advantages of annual revaluations in terms of fairness, ensuring that RVs were always reasonably close to the valuation date. It is also suggested that three-year revaluations would lead to greater stability compared with a five-year cycle, because there would probably be fewer significant changes in rateable value over the shorter period.
G58. However, the IPD data in appendix 3 show that there can still be considerable turbulence in rental levels over a three-year period. There were increases of over 60% in 1988 and 1989 and 45% in 1990, compared with three years earlier, and falls of 17% in 1992, 26% in 1993 and 20% in 1994. These are the extremes and in some other years a three-year cycle could produce a more stable outcome, but there is no guarantee that it will do so at every revaluation.
G59. In the worst case, a three-year cycle could have produced an increase of 39% in 1987, followed by a 45% rise in 1990, then a fall of 26% in 1993. This hardly provides stability. No three-year cycle could have avoided the four successive years of big increases in 1987-90 and the four successive years of big falls in 1992-95. There are also significant variances between sectors, with nothing less than 15% in any year between 1985 and 1995 and closer to 30% in most of those years. Only in 1996 and 1997 was there a variance as low as 10%.
G60. The VOA data in appendix 3 show that there are also significant variances across the English regions on a three year cycle. While there are periods of relative stability in any one region and sector, there are also periods of great change. And the periods of stability are not concurrent in all locations and sectors. As a result, the lowest variance is 44% in autumn 1995. In most periods the variance exceeds 60%, reaching 160% in autumn 1991.
G61. This suggests that three-year cycles will not prevent significant changes in rate bills. Even if changes in bills may be less overall than with longer cycles, they would still cause the same kinds of problems for those ratepayers affected by them. It is therefore unlikely that three-year cycles would eliminate the need for transitional relief.
G62. Both the IPD and VOA data show that there is generally less turbulence with a three-year cycle than with a five-year cycle. However, there is not significantly less, as the changes and variances in several three-year periods exceed those in some other five-year periods. So there is no guarantee that a three-year revaluation will always produce less turbulence than a five-year revaluation.
G63. Reducing the period between revaluations will in itself reduce stability and certainty, whatever their outcome. There will only be two years of relative stability before the next revaluation comes along, creating the next round of uncertainty and changes in rate bills.
Five-yearly revaluations
G64. The terms of reference for this review state its aim to improve certainty and stability, while maintaining fairness. The current five-year cycle is clearly not ideal in attaining these objectives. The 1995 and 2000 revaluations have produced many big shifts in rate bills. The revaluation process itself has caused considerable uncertainty for ratepayers. By the end of a five-year rating list values are almost seven years out of date. However, 52 responses to the consultation favoured retaining the current cycle.
G65. While a five-year cycle may not be ideal, there is no certainty that reducing it would deliver greater stability, in terms of significantly less turbulence in rate bills. Without that, more frequent revaluations would simply increase the frequency of the uncertainty and instability caused by any revaluation. The only advantage of more frequent revaluations would be the increased fairness of rateable values more in line with actual rental levels.
G66. Five-year cycles have two unique advantages. The first is that most commercial rents are negotiated on the basis of a five-year review period. The second is familiarity. Ratepayers are becoming more accustomed to the five year cycle with each revaluation following the 17 year gap that preceded the introduction of NNDR in 1990. It may be that the level of anticipation of the revaluation is now somewhat higher than previously and that this has helped provide such a good reception for Revaluation 2000.
G67. The IPD data for a five-year cycle in 1998 shows an increase of 19% for the UK as a whole, compared with 1993 values. –This period is directly comparable with the valuations for Revaluation 2000 in England, Wales and Scotland. The IPD UK figure of 19% compares reasonably closely with the actual revaluation effects of 25% in England, 13 % in Wales and 13% in Scotland over the same period. There was no revaluation in Northern Ireland in 2000. The variance between sectors on IPD data was 16%, the lowest of any period. The VOA data shows a regional variance of 91% in spring 1998, again one of the lowest.
G68. As noted above, both the IPD and VOA data show that, while there are generally more changes than for a three-year cycle, there can be periods where there is less change on a five-year cycle. For example, on the IPD data, the five-year value changes for 1998 quoted above would have been exceeded on a three year cycle in 1998, 1999 and every year from 1986 to 1990, while the variances would have been greater every year from 1986 to 1994. On the VOA data, the five-year variance for spring 1998 would have been exceeded on a three-year cycle in every period from spring 1991 to spring 1993, as well as in spring 1994 and spring 1999.
Ten-yearly revaluations
G69. One way to reduce the uncertainty and instability caused by any revaluation is to hold revaluations less frequently. This would give a greater period of certainty and stability. It would also allow more time for any transitional relief to unwind. If there were a ten-year revaluation cycle, there could be a TR scheme that unwound over ten years, rather than three or five, allowing smaller steps towards the full liability, which may be more manageable for those facing the biggest increases. Alternatively, if TR unwound over say three years, there would then be seven years of complete certainty and stability, with everyone paying their full liability, before the next revaluation.
G70. However, there was little support amongst respondents to the consultation paper for a longer revaluation cycle. Only 7 out of 114 favoured a seven or ten year cycle. The review group did not give explicit consideration to a seven year cycle, as the same arguments about balancing stability and certainty against fairness apply as for other cycles. As with three-year cycles, there is unlikely to be a significant difference compared with a five-year cycle.
G71. There was significant concern that a ten-year cycle would allow rateable values to become very much out of step with market rents and lead to significant step changes in value come revaluation. This could cause misgivings amongst the public about the fairness of the tax, particularly if ratepayers have to wait for a decade before benefiting from any correction in rateable values to reflect the rents they are paying for their properties.
G72. For example, the IPD data shows a 5% change in values in 1998 compared with 1988 - the valuation dates for the 1990 and 2000 revaluations. Had a 10-year cycle been employed the turbulence of the 1995 revaluation would have been avoided and that of the 2000 revaluation may have been reduced. We know that, in some parts of the country, values fell at the 1995 revaluation and have risen again at the 2000 revaluation and are now closer to their 1990 values. While not having the 1995 revaluation would have reduced instability in rate bills, it would have been unfair on those whose values were lower on the 1995 rating lists, reflecting their reduced values at the time.
G73. The 5% increase on a ten-year cycle compares with a 19% increase in the five years before 1998. In most other years a ten-year cycle produces bigger changes - as much as 142% in 1990. Even so, the biggest changes in value on a five-year cycle, in 1988, 1989 and 1990, are only exceeded by those on a ten-year cycle in 1990 and 1991. So a ten-year cycle does not necessarily produce higher changes in values than a five-year cycle.
G74. The variances between sectors are generally greater for a ten-year cycle. In 1998 there was a 55% variance, as opposed to 16% for a five-year cycle. This suggests that a ten-year cycle may have produced greater increases and decreases in bills than a five-year cycle - retail rental values rose by 36% but those for offices fell by 19% - even though the overall change in values was less on a ten-year cycle. The variance for a ten-year cycle is around 50% to 60% in every year back to 1992 (rising to 67% in 1993). A five-year cycle produced sectoral variances of up to 44%. This suggests that a ten-year cycle would probably lead to more changes in rate bills, though in many years these may not be significantly more than those produced by a five-year cycle.
G75. The VOA data only shows four ten year periods, ending between spring 1998 and autumn 1999. Again, these show substantial variances, but are exceeded by the greatest variances on a five-year cycle. The five-year figures for 1993 show greater variance than the ten-year figures for 1998. Both are compared with a base year of 1988. Again, this suggests that the effect of a ten-year cycle on bills is likely to be more volatile than a five-year cycle in most years, though it may not be much more volatile. This effect may be offset by the decreased frequency of revaluations. In some years a ten-year cycle could have less of an effect on bills than a five-year cycle. This reflects the interaction of any revaluation cycle with the peaks and troughs of the property cycle.
G76. The advantages of a ten-year cycle in terms of a longer period of stability and certainty between revaluations must be considered against the greater turbulence likely to be caused when revaluation comes, causing more instability and uncertainty. There would be a bigger hit, less often, but it need not be much bigger than with a five year cycle - in some years it might even be less. The effect on bills could be mitigated in both cases by transitional relief.
G77. However, a ten-year cycle would be comparatively less fair than a shorter cycle, as by the end of it rates bills could be based on twelve-year old valuations. How unfair this would be depends on how much the relativities between values across the country had changed over that time. It would still be significantly less than the 17 years between the 1973 and 1990 revaluations. But it is a major concern for ratepayers.
Conclusions
G78. There was no clear consensus on any benefits from moving from the current five-yearly revaluation cycle. Revaluations are necessary to maintain a reasonable degree of accuracy in valuations, to provide fairness and acceptability, given the system of precise individual valuations, but when they happen they cause instability and uncertainty. In essence the arguments were clear:
- the shorter the valuation cycle the fairer the system is, because the values on which liability is based are more up to date, BUT the system is less stable and certain for ratepayers as values change more often;
- the longer the cycle the less fair it is for ratepayers, as values becomes less accurate over time and may reach a point where they are no longer acceptable, BUT there is greater certainty and stability over the amounts they will have to pay for a longer period;
- more frequent revaluations will cost more than less frequent revaluations;
- more frequent revaluations may reduce the number of appeals at each revaluation, but may increase the total over time.
G79. Any decision on revaluation cycles, whether to make them shorter, longer or leave them unchanged at five years, depends on the relative importance given to these factors. It is also interdependent with decisions on other matters considered in this report, in particular whether there is a transitional relief scheme.
[ Annex G Section2 ] [ Contents ] [ Annex G Section4 ]
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