Response to government consultation on business rate retention

Date: 16 June 2017

I. Introduction

This is a response by David Phillips, an Associate Director at the Institute for Fiscal Studies (IFS), to the government consultation “100% business rates retention: further consultation on the design of the reformed system”. The views and opinions expressed here are those of the author only. The IFS has no corporate views.

Alongside other IFS researchers, David is currently building a dataset and quantitative model of local government finance. The first component involves looking at how revenues diverged in the past, and the extent to which these revenue changes correlated (either positively or negatively) with changes relative spending needs as assessed by RNF scores. The second component will involve projecting forward revenue and spending needs (and uncertainty around these projections). The third component is a set of code to simulate the impact of different variants of the BRRS (and other LG finance system parameters) on the degree of divergence between relative funding and relative spending needs, and the strength of fiscal incentives for revenue growth, given these projections and user-defined 'scenarios' for local revenues and spending needs.  This model will therefore allow quantitative analysis of the issues raised by some of the consultation questions. At this stage though, it is possible to provide qualitative responses to the consultation questions.

II. Responses to consultation questions

  • The approach to resets

The retention of 100% of the growth (or decline, subject to a safety net) in business rates revenues provides councils with the incentive to grow their rates revenues (e.g. by promoting economic and property development). But if such growth could be retained in full indefinitely, we could see very large divergences in revenues and spending needs open up. Resets of the system are designed to stop this from happening, redistributing revenues from areas that have grown to areas that have fallen behind. However, in doing so, they blunt the incentive to grow revenues. A partial reset can be seen as an attempt to trade-off the objectives of "equalisation" on the one hand and "incentives" on the other:  councils can retain some (but not all) of the growth beyond the end of the period.

I and my team hope to engage with DCLG (and other stakeholders) on this issue using our model of local government finance, currently in development. At this stage I would make the following observations.

A) The appropriate length of time between reset and % reset to adopt should depend on:

  • How large divergences in revenues may grow to be in the medium and long term (i.e. do revenue divergences grow between councils or do revenues revert to mean);
  • The extent to which one believes divergence in revenues (and spending needs) reflects impact of council decision making versus wider economic trends. If latter, the "incentive" is not particularly relevant: it is just risk;
  • The tolerance for divergence in revenues (and resulting impacts on service provision) by policymakers and the public.

It is not clear how such considerations have influenced the choice of a 5 year reset period, or how they will influence the choice of the proportion of growth to be retained beyond the 5 year reset period. Indeed, the proportion to be retained appears to be something of a "residual", being defined as "what’s left" after ensuring those who have seen a real-terms reduction in their revenues receive at least the baseline funding level (indexed to inflation). This is not particularly satisfactory, but could be avoided if:

i) councils were allowed to retain a proportion of "losses" relative to baseline funding beyond the 5 year period;

ii) the system were not required to be fully self-funding (which it seems unlikely to be in the long term, given rising demand for services like adult social care).

B) As IFS researchers have previously commented (Adam and Miller, 2014; Amin Smith et al, 2016) there would be benefits of using a rolling reset as opposed to a fixed-reset: it provides certainty of retained rates growth from a given development for a fixed period of years, rather than up to a fixed date which, if close to, could actually encourage delays to development to push it into the next five year period (although we note partial resets lessen the severity of this problem).

C) As highlighted in working group meetings between DCLG and the LGA, there may also be benefits in undertaking resets of spending needs more or less frequently than for revenues. Whether it should be more or less frequent would depend on: The extent to which divergences may be greater or smaller for spending needs as opposed to revenues (and again issues like whether spending needs exhibit 'reversion to mean');

  • Whether it is felt councils are able to influence spending needs more or less than business rates revenues (and whether this influence is 'good' - such as tackling the drivers of ill-health, poverty, etc - or 'bad' - making the area less attractive for people with ill health or low incomes by, for instance, cutting back services).

Also of relevance is the extent to which changes in spending needs and revenues are correlated with each other: if increases in revenues are associated with declines in spending need (and vice versa), this increases the potential for divergence (which may make more frequent or fuller resets more desirable) but also means that resetting one (e.g. spending needs) may weaken incentives to take action on the other (e.g. revenues).

While it is recognised the there was a desire among most councils for resetting spending needs and revenue retention over the same cycle, and there are benefits of such an approach in terms of ‘simplicity’, it is hoped that any final decision is informed by quantitative modelling.

D) It is also worth noting that 'transitional arrangements' on the spending needs side act to slow down the adjustment to a reset (making it even more 'partial', including for "losers" during the preceding period). Its use should be kept to a minimum and detailed modelling, comparing the ability of more frequent resets, use of 'smoothed' needs data, and transitional arrangements, to avoid overly sharp changes in funding at resets should be undertaken.

E) We now turn to an issue not mentioned in the consultation: the uprating of tariffs and top-ups (and of baseline funding) between resets. It is implicitly assumed that this should be in line with inflation. As set out in Amin Smith et al (2016) such an approach can lead to divergences in relative funding (compared to national pooling of revenues) even if business rates revenues grow at the same proportional rate in all council areas. If instead these parameters were increased in line with (forecast) growth in national business rates revenues, this issue would not arise. As councils would still bear 100% of the marginal gain/loss from changes in business rates revenues, this would not affect their incentive to grow revenues.

F) Lastly, the consultation also omits references to how resets relate to council tax bases. Council tax will make up around half of councils’ revenues, although this will vary significantly across areas. Furthermore, council tax bases may change differentially as new housing units of different types are built (e.g. small flats versus large houses) and become home to different types of people (e.g. single low-income pensioners, versus high-earning young couples). Differential changes in council tax bases could impact councils’ ability to fund services but there is no discussion of how council tax bases will be taken into account at the initial or subsequent resets. This seems to be an important omission, particularly as the balance between revenues from council tax and business rates – the multiplier for which will be capped at CPI – may change over time.

2. Defining business rates growth

It would seem sensible for the definition of growth to be consistent with the measure used to uprate tariffs and top-ups (e.g. inflation or national business rates growth), and to assess growth against the business rates baseline.

Whether to use a single point of time or an average over several years should depend on the time series behaviour of business rates: if it highly volatile and/or mean-reverting, use of an average would seem appropriate, but if it less volatile and increasingly diverges, single points may be more appropriate. However, it is worth noting that if a rolling reset were utilised (as opposed to fixed cut-offs) this issue would not arise in the same way.

3. Pooling arrangements

Examination of pooling agreements under 50% scheme shows that avoidance of levy was a key motivator for pools. Without levies, in the absence of new incentives, pooling could only reduce aggregate retained revenues at the level of the pool (through loss of safety net payments) so there would appear to be a strong risk of pooling arrangements - and potential for risk sharing - petering out.

Incentives suggested include the ability to retain more growth following resets and more generous safety net thresholds. Together these would provide incentives to pool for pooled areas that envisage doing well out of rates retention (and so would gain from retaining more growth) and (at risk of) doing poorly out of it (and so would potentially gain from the more generous safety net).  But they would still not provide councils to agree to pool with a particular authority thought to have a high risk of falls in rates revenues.

With this in mind there is a rationale for giving DCLG the power to mandate pooling (and pooling groups) - although it is clearly a diminution of councils’ autonomy. It is important to ask whether DCLG will have the power to mandate the allocation of revenues within a pool (e.g. the percentage of 'insurance' within a pool, etc.), as well as the creation of the pool.

Parameters chosen for each of the incentives, as well as local growth zones, would need to be carefully considered in the context of the sustainability of the system as a whole. In particular the more generous the incentive, the stronger the incentive will be for councils to pool, but the less revenue will be available to councils not in a pool.

There may be particular issues in relation to local growth zones. Greater retention of revenues within these zones may skew councils’ incentives to promote development in the particular geographic areas covered by the zone at the expense of other areas. If agglomeration effects are important, such an incentive may be useful. Alternatively, there is the possibility that such skewed incentives could even reduce overall development and aggregate business rates revenues if councils promote development in these areas much more than in other areas, or reduce rates to attract occupiers to these locations (from elsewhere in their locales).

Councils will also have a strong incentive to select areas that they expect to perform well in terms of business rates revenues as their growth zones, meaning achieved growth may largely reflect what would have happened anyway, rather than the incentive provided by the growth zone. Furthermore if a substantial proportion of all revenue growth took place in these zones, the quantum of revenues available for redistribution may make the trade-off between compensating ‘losers’ and allowing retention of some of the growth among ‘winners’ for areas outside growth zones particularly challenging. 

Given these issues, it will be important to undertake comprehensive analysis of the potential effects of different options for these incentives. The rationale for local growth zones – as opposed to allowing pooling councils to retain more of the growth beyond resets in the entirety of their areas – should also be considered carefully.

4. Centrally-managing appeals risk

Recent research at the IFS (Amin Smith and Phillips, 2017), demonstrated that there has been significant variation in the provisions made for appeal costs and the charges against these provisions by councils. This suggests that there is significant risk at the local level from appeals diverging from the forecast national rate of appeals – which councils currently retain as their own ‘top slice’ to fund appeals costs. Given that valuation errors are outside councils’ control, insulating councils from this significant revenue risk seems sensible.

As we understand it, government’s plans involve retaining the ‘top slice’ centrally and then redistributing it to authorities according to need as successful backdated appeals arise – rather than distributing it according to business rates baselines up front. To do this, the following change will need to be made to the operation of the BRRS:

  • Rather than basing councils’ ‘business rates baselines’ on how much they would raise after allowing for an estimate of the average cost of appeals nationally, ‘business rates baselines’ will be based on how much would be raised before allowing for appeals.

Because the amount to be distributed according to need (the ‘baseline funding’) would still be based on the lower national revenues after accounting for expected appeals, this system means that, across England as a whole, councils are assumed to be able to (initially) raise more business rates revenues than they need. In other words, the approach generates a net tariff on councils’ business rates income. This net tariff can then be used by central government to centrally fund the cost of appeals.

A similar change would occur when a revaluation takes place. Rather than basing the new tariffs and top-ups on the impact of the revaluation on councils’ business rates income using the multiplier before it is upwardly adjusted to raise revenue to pay for appeals (e.g. 0.436 in the 2017 revaluation), one would use the multiplier after that adjustment (0.457). Because baseline funding would not be upwardly adjusted at the same time, this would also generate a net tariff on councils’ business rates income that central government could use to centrally fund the cost of appeals.

A question then arises of what will happen, given that the government is highly unlikely to fully accurately predict the cost of appeals. If its estimate of appeals costs is too low, the net tariff generated will be too small to pay the costs (and vice versa). If this is the case, central government could make up the difference (or retain the difference if it overestimated appeals costs). Alternatively, it could retrospectively alter the top slice applied to baseline funding, making the local government sector as a whole bear the risk of appeals across all councils coming in over or under initial forecasts.

The first of these approaches generally seems preferable, as it would prevent councils (as a group) seeing their income reduced or increased compared with initial expectations, if the government initially under- or over-estimated the cost of successful appeals – errors that councils would not have been responsible for, and a risk over which they had no control. Top-slicing baseline funding to cover an underestimate of likely appeals would also, in general, have a greater impact on relatively poorer councils with smaller council tax bases which, under 100% retention, will rely on the BRRS (including top-ups) for more of their overall budget.

DCLG should also continue to liaise with the VOA on ways of making the information available on ratings lists as useful as possible for this exercise (e.g. splitting out changes in valuations due to challenges of initial valuations versus changes in properties / circumstances).

5. Tier splits

The consultation rightly highlights that tier-splits (40% for districts, 9 or 10% for counties) in two-tier areas mean that:

  • Districts presently are ‘highly geared’ and therefore have strong financial incentives to grow their business rates revenues (a 1% growth in revenue translates into a much more than 1% increase in retained rates revenues) but also face greater revenue risk (subject to the safety net).
  • Counties presently have ‘low gearing’ and therefore have weaker financial incentives to grow their business rates revenues (a 1% growth in revenue translates into a much less than 1% increase in retained rates revenues) but also are insulated from revenue risk.

It does not appear to recognise that these tier splits also mean that, in the long-run, if rates revenues grow (fall) in real terms then the lowly-geared counties capture a smaller (larger) proportion of overall business rates revenues. Given that counties are responsible for a service area – adult social care – that may be expected to face particularly strong spending pressures in the coming years, this may present challenges for the sustainability of funding of county-level services.

One approach to this issue would be to increase tier shares for counties relative to districts (unless business rates revenues were thought to be on a downward trajectory). Doing so would also:

  • Increase incentives for revenue growth for counties, relative to districts (which may or may not be seen as desirable given existing and future services/powers related to growth);
  • Increase revenue risk for counties, relative to districts (which again may or may not be seen as desirable).

Rather than adjusting tier shares to ensure counties capture a larger proportion of business rates revenues in the long term (than they would otherwise), an alternative would be to uprate tariffs, top-ups and baseline funding levels at a rate greater than inflation: for instance, forecast national growth in business rates revenues (if greater than inflation). Uprating tariffs, top-ups and baseline funding levels in line with national growth in rates revenues, would not affect incentives/risks at the margin in the same way as changing tier shares, and would lead to counties’ retained rates revenues keeping pace with national rates revenues growth, in expectation (rather than falling behind if revenues grow in real terms).

Overall though the approach for choosing tier shares should be based on an understanding of counties and districts abilities to act on incentives and bear revenue risk, and an understanding of how tier-shares (and gearing) may affect the distribution of funding between tiers in the long-term (and how this interacts with other features of the system such as rules around uprating and resets). Quantitative modelling would be useful in this regard.

6. Safety net

The choice of safety net threshold is a trade-off between providing insurance to councils against falls in their revenues (that may or may not reflect factors outside their control), and providing incentives to councils to promote revenue growth (which safety nets blunt, as those receiving safety-net payments retain none of the revenue growth until they exit safety net payments).

I will make four points in relation to safety net payments:

a. The likely scale of safety net payments will depend on how many councils have high levels of gearing (highly geared councils are more likely to require safety net payments as small falls in revenues can lead to large falls in retained revenues), and how many councils pool (as pools both reduce gearing and allow for insurance within the pool without recourse to a safety payment). Safety net thresholds and top-slices should be set with reference to these ‘risk’ factors

b. If one wanted to continue to provide councils in receipt of safety net payments with some incentive to (marginally) boost their revenues, rather than providing full coverage of losses below a threshold, it would be possible to provide partial coverage of losses below a (potentially higher) threshold (perhaps on a sliding-scale so that more coverage is provided for larger losses)

c. There is no mention of how safety net thresholds relate to the proportion of funding that is obtained from the BRRS (relative to council tax). Areas with high spending needs / low council tax bases will receive more of their funding from the BRRS so a given safety net threshold would imply a larger fall in their overall revenues (including from council tax) in areas with low spending needs / high council tax bases. There may be merit in setting safety net thresholds in relation to a measure of funding that accounts for ability to raise revenues via council tax, rather than simply as a fraction of BRRS baseline funding.

d. A decision will need to be taken as to what to do if the top-slice made to fund safety net payments is too small / too large. Rather than seek to reconcile in subsequent years, it may be simpler for central government to bear this risk, although this could incentivise central government to over-estimate the top-slice required.

7. The central list

The central list currently operates as a home for properties that by their nature are difficult to value at the level of an individual local authority (or indeed value at all): rail, communications and utilities networks, for instance.

The consultation states that an updated statement of policy on “which properties and ratepayers should be assessed on the central list” will be provided ahead of the introduction of the 100% BRRS.

As discussed in the previous consultation, it should be considered whether the central list should also be used as a home for those properties that have high levels of risk that are outside councils’ control and where central government policy or global economic conditions may have more of an influence (power stations, airports, oil refineries, etc). Centralising these properties will remove the risk of significant revenue losses (subject to the safety net) and windfall gains for councils, although it will also remove local fiscal incentives to support such activities (which may involve local costs – traffic, pollution, noise).

In addition to considerations of risk, it may also be worthwhile considering the central list ‘boundaries’ in relation to the level at which planning permission will be granted (e.g. whether by the local authority, or as part of national infrastructure planning). The national infrastructure commission is particularly interested in where powers and fiscal incentives lie in relation to nationally important infrastructure and may be worth liaising with.

III. References

Adam and H. Miller, ‘Business rates’, in C. Emmerson, P. Johnson and H. Miller, The IFS Green Budget: February 2014, IFS Report R91, 2014, available at https://www.ifs.org.uk/publications/7072

Amin-Smith, D. Phillips, P. Simpson, D. Eiser and M. Trickey, ‘A Time of Revolution? British Local Government Finance in the 2010s’, IFS Report R121, 2016, available at https://www.ifs.org.uk/publications/8705.

N. Amin-Smith and D. Phillips, ‘The Business Rates Revaluation, Appeals and Local Revenue Retention’, IFS Briefing Note BN193, 2017, available at: https://www.ifs.org.uk/publications/8962.