As discussed in our how to evaluate series, one way to ensure robust evaluation is to make better use of secondary data. This can be a cheaper and less burdensome way to find out about both the ‘treatment’ and ‘control’ groups, since this data is already being collected for other purposes.
When tracking outcomes related to firms an important secondary source of data is the Inter-Departmental Business Register (available here). Researchers can use the IDBR to access data on individual firms across multiple years (i.e. panel data). The ONS VML or the ESRC SDS (here) are both user-friendly services that facilitate access to the IBDR. BEIS even have a quarterly version of the IDBR which, in principle, could allow quick turnaround on assessments of interventions aimed at helping specific firms.
Making use of this data for policy development will often require that locally held data (for example, on firms that are participating in growth hub schemes) can be linked to the IDBR. This is not always a straightforward process, but here are some suggestions on how to improve matching rates, increasing the percentage of firms that can be successfully linked to their IDBR records:
- Company Registration Number (CRN) is the main piece of information that gets a decent match to IDBR. CRN is the most practical to collect and sufficient to get a good, but certainly not complete, match rate.
- VAT registration numbers can increase matching rates (by around 5 percentage points).
- If you want to collect data on enterprises that are not limited companies – possibly to match to other data sources that could become available – then it is helpful to collect VAT, PAYE or Tax ID (for self-employed non-employers).
- Other organisations may have complex structures making a match less straightforward. Here, accurately recorded name and address can help. When you collect addresses, try to find out if it’s the HQ or a local unit of a larger firm. And make sure you collect the postcode!
- Details may not be accurately collected – best to check data as early as possible, preferably with a system of live validation.
- Finally, it is helpful where possible to match to multiple extracts of the IDBR for best accuracy.
Linking local data to the IDBR in these ways can help produce better evaluations of local growth policy, and so strengthen the evidence base on what works and doesn’t work.
As always, we are happy to provide further advice, so please feel free to contact us to discuss further.