Small Business Bonus Scheme: evaluation

8. Discussion and recommendations

We have undertaken an evaluation of the Small Business Bonus Scheme (SBBS) using three methods: 1) a descriptive analysis of the non-domestic rates data; 2) a survey of the small business base in Scotland; and 3) an econometric analysis of business outcomes using modern econometric techniques suited to identifying the effect of the SBBS.

We documented the limitations that we faced in conducting the research due to:

1) the nature of the data and our uncertainty in the accuracy of the database of businesses, which affected both our ability to survey businesses and our ability to match businesses with administrative data;

2) the 'bunching' of non-domestic properties around policy thresholds which violates some of the assumptions underlying our econometric analysis; and

3) the limitations imposed by the Covid-19 pandemic.

Notwithstanding these limitations, we have undertaken our assessment of the policy and have documented our findings from the various aspects of our research. These are summarised in the introduction and detailed throughout the report.

A key finding from our analysis of both the administrative data and the self-reported data collected in our survey of businesses is that we cannot find any statistical evidence to conclude that the SBBS helps businesses that receive 100% relief achieve any better business outcomes (in terms of turnover; employment; GVA) than those in receipt of only 25% relief but with a similar rateable value. However, as we have documented throughout, the limitations we faced mean this should be interpreted as being that there is no evidence of an effect rather than there being evidence of the SBBS having no effect.

The responses from our survey are suggestive, however, that the SBBS supports businesses in various ways: there is some evidence that the SBBS has a marginal effect on the likelihood of paying the Real Living Wage; SBBS recipients are slightly more likely to have reported increases in investment expenditures; and businesses that receive full relief report more of an effect of the SBBS on their businesses than those that receive partial relief.

What has become clear, however, is that there is a high degree of variability between businesses in narrow rateable value bands. This is evident both from administrative data, the information provided in our survey on employment and turnover, and businesses responses to the effect of a change in costs. This means that businesses within the same narrow rateable value band can vary hugely in size and presumably, therefore, in terms of need for business support. Yet, rateable value is the primary criterion on which SBBS eligibility is determined.

Part of this may be attributable to the difficulties associated with constructing a database of businesses from the data. Moreover, part may be due to recorded (in the data) or reported (in the survey) rateable value only capturing the Scottish properties of businesses and they have interests elsewhere (22% of eligible claimers reported to have properties outside Scotland), so the full size of the business is not being captured by Scottish rateable value. But we suspect that even once these issues are accounted for, substantial variability among businesses judged to be similar according to rateable value will remain, which brings into question whether rateable value is a reasonable proxy for the need for business support, and so whether it should be used as the basis for SBBS eligibility.

We conclude our evaluation with a series of recommendations for the Scottish Government to consider. As an independent research institute, we do not advocate specific policy choices, so the recommendations and reflections here are designed to help support a broad conversation on the future direction of the SBBS and non-domestic rates policy more generally.

1. Firstly, as we highlighted throughout this evaluation, it is exceptionally difficult to reach any definitive conclusions about the impact of the SBBS on the business base in Scotland. Challenges around data quality, identification of businesses and policy implementation (the bunching of businesses around policy thresholds) make it almost impossible for any systematic assessment of the impact of the SBBS to be undertaken on key issues such as employment, investment or business growth. Policymakers and the business community will no doubt agree that it is important to be able to draw robust conclusions about the efficacy of policy using the best available research methods to help inform future policy design. This is true for an important policy such as the SBBS, particularly given the level of investment undertaken over the now thirteen years in which it has been in operation – from £93 million in 2008 (in 2020 prices) to £279 million in 2020. We recommend therefore that the Scottish Government reflects upon how it builds in information gathering, data collection and evaluation into its policy development process on all aspects of its business support policy agenda from the outset.

2. Following on from this, one of our particular recommendations is that the Scottish Government begins to collect new information that will make a more robust assessment of the SBBS (and potentially other reliefs) possible in the future. What is required to undertake an effective evaluation of the SBBS is an accurate database of businesses (as opposed to a database of properties) and associated business outcomes. Our suggestions include:

a. The creation of a digital registry of businesses in Scotland, maintained by the Scottish Government, requiring each business operating in Scotland to maintain a digital registry entry providing its contact details (including email address) and detailing all properties the business owns, rents or has use of (both within and outside Scotland). This would be in essence an investment to create a sub-set of the Inter-Departmental Business Register in Scotland, with links between business level and local unit (or property) level, in order to link properties properly to business units. A separate business register is currently in place for Northern Ireland, but the current Inter-Departmental Business Register is for Great Britain as a whole.

b. Ensuring this digital registry is periodically updated.

c. Ensuring the database of businesses in the digital registry contains the appropriate identifiers so that businesses can be accurately matched to their record in administrative data sets.

d. Maintaining a database of business outcomes of interest (e.g., turnover, employment, investment) for the business base in Scotland (including micro, small and large businesses), updated annually, that can be used to evaluate the efficacy of the SBBS. Ideally this would be statutory reporting of data by the whole population of businesses (some of this data, but by no means all, could be taken from existing government data sources). Failing that, the digital registry can be used as the sampling frame to conduct an annual (statutory) survey of Scottish businesses outcomes in the year. This would replicate both the approach that the Northern Ireland Statistics and Research Agency and the ONS take for Northern Ireland and Great Britain respectively. In some cases, such as Scottish single site businesses, the datasets used by the ONS can be utilised in the Scottish context (such as the use of HMRC VAT turnover data to update turnover information).

e. Requiring businesses to report appropriate data on other business activity outside Scotland (especially, but not limited to, those receiving non-domestic rates relief) to allow a better assessment of the need for business support within Scotland.

f. Ensuring guidelines for data collection and management are consistent across local authorities and assessors.

3. Using this data the Scottish Government should undertake regular and comprehensive assessments of the SBBS, alongside other business support policies. This will enable the impact of the SBBS to be reviewed regularly, both against its own objectives and alongside other policy initiatives. Furthermore, robust evaluation of this data will facilitate developing a better understanding of the determinants of the efficacy of the SBBS. To do this will require a step change in the approach to policy design and data collection, but will enable policies like the SBBS to be compared in relation to other government business support schemes to identify the most effective package policy interventions for Scotland.

4. Fourth, we find 'bunching' of non-domestic properties around SBBS eligibility thresholds. The location of this bunching changed over time as the policy thresholds changed, and was partly due to much higher rates of appeals among those valued just above the 100% threshold in the 2016/17 revaluation. We recommend that the Scottish Government explore with the independent Scottish Assessors why this occurred.

5. Fifth, we find that rateable value is a poor measure by which to classify businesses as 'small', and therefore in need of support. Businesses within narrow rateable value bands are substantially different in terms of business outcomes (even when attention is restricted to a particular class of business). For example, our evaluation has found that businesses with large turnovers but operating from low rateable value properties are eligible for SBBS relief. It is not clear that this is the objective of the policy in supporting 'small businesses'. Once it has implemented an effective data collection programme to allow an effective evaluation of the SBBS, the Scottish Government may wish to reflect upon the aims and objectives of the scheme and how well-targeted to achieving these it has become over the years. Of course, alongside this, the government is likely to consider how large turnover businesses may be better supported or incentivised by means other than the SBBS. As set out above, this would require the development of explicit policy objectives, which would in turn allow for targets against which the policy's efficacy can be assessed.