Exploration of data-matching as an evaluation technique: merging of ‘Designated Relationship Managed’ companies into the ONS Annual Respondents Database and econometric analysis

The Account and Client Management programme is a Direct Relationship Management (DRM) intervention, under which Scottish Enterprise (SE) provides support to companies that are considered to be capable of benefiting from a high level of attention. This report considers the feasibility for undertaking an econometric analysis of DRM support by combining SE records on Account and Client Managed businesses with business performance data from the ONS’s Annual Respondents Database (ARD). The evaluation aims to help determine the causal impact of DRM interventions on individual business performance and the Scottish economy.

The DRM data supplied by SE was reviewed and matched into the ONS ARD and the characteristics of the linked dataset were then analysed. The performance of the DRM firms was measured against a control group of firms drawn from the North East and North West regions of England, to test if DRM status made a difference.

Various data issues had to be considered that ultimately lead to a rather small sample being available for econometric analysis. However, it is concluded that this analysis is providing SE with useful information on whether DRM status made a difference. The results suggest that Scottish DRM firms did not do better than the control group of similar firms from the North East and North West of England, with the statistical evidence pointing to them having done worse in terms of total factor productivity (TFP). There is some weak evidence that, before being assisted, DRM firms had higher productivity; in the year when assistance started productivity significantly declined by around 9% with further, although insignificant, falls until years 5 and 6. From the 6th year onwards there was a significant overall decline in TFP of nearly 16% in Scottish DRM firms in comparison with the control group of similar firms in the North East. With regard to the control group from the North West, the results show that there is again some weak evidence that several years before being assisted, Scottish DRM plants had higher productivity, which fell in the year assistance started. After assistance, there were further, albeit insignificant, falls. Again, from the 6th year onwards there was a significant overall decline in TFP of 33% in Scottish DRM firms in comparison with the control group of similar firms in the North West.

Recommendations are made on what needs to be taken into consideration if this type of analysis is to be repeated in the future, when more and better data are available.