Data analysis
5.107 Some of the arguments advanced in opposition to the proposal were based on assumptions about the quality of the SARS that are currently filed. In order to establish how many reports were being lodged based on suspicion without objective grounds in support, we analysed a sample of 563 authorised disclosures. Of the 563 that we considered, only 295 were based on one or more objective grounds. This amounted to 52.4% of the sample. 47.6% of our data sample did not demonstrate reasonable grounds for suspicion. This represents a substantial proportion of authorised disclosures which are lodged without objective grounds in support.
Analysis
5.108 The suggestion that a change in the test would create a barrier to prosecution is not borne out by the practical reality of how these types of cases are prosecuted. The Crown Prosecution Service have confirmed in further discussion during the consultation period that the key challenge is demonstrating to a jury that objectively, there were reasonable grounds for a defendant to suspect. From this, a jury is likely to infer that the defendant must have been suspicious. From this perspective, a change to the test would not make prosecutions more challenging or introduce a new barrier.
5.109 There would be two possible consequences to amending the threshold to one based on reasonable grounds. First it could result in the same volume of SARs being lodged, simply with better quality intelligence and more raw data for law enforcement agencies to examine and utilise. Secondly, it could also have an impact on the quantity of disclosures being made if reporters felt that they did not have sufficient material in support of their suspicion to meet the higher threshold. However, our broad conclusion from analysing authorised disclosures was that, in the majority of cases, reporters had access to the information but had chosen not to provide it or had omitted it for other unknown reasons.
5.110 Having considered the many valuable points made by consultees and the data analysis we conducted, we consider that answering this question involves a difficult balancing exercise. It must not be forgotten that none of the responses refer to the impact on those who are the subject of disclosures. As we identified in our Consultation Paper, a balancing act is essential as individuals and businesses may suffer anything from temporary inconvenience to severe economic hardship as a result of being unable to access their funds. A complete analysis needs to consider whether resources expended in the gathering of intelligence are proportionate bearing in mind other impacts. One consultee, a barrister named David Lonsdale, had both his business and his personal accounts frozen. He described the impact it had on him:
I think there are far too many trigger-happy irresponsible reports made by the banks in relation to many of their customers and these coupled with the freezing of accounts do untold damage to the lives of many innocent people. The practice also wastes a great deal of the time and resources of the NCA.
The second bad aspect of the current law is that it is simply inhumane to freeze a person's account for up to eight working days without giving him access to any money at all. (In my case I had £20 in cash in my flat and no food and there was no one in London who could help me. I was left absolutely desperate.) But the banks think they must do this pending consent from the NCA or the passage of 8 working days. If the NCA received a report in a coherent prescribed form that related only to really serious crime, then there is no reason why it should not be able to decide within 48 hours whether consent could be given or not.
5.111 Whilst we concede the possibility that a change to a test based on reasonable grounds to suspect as the trigger threshold for reporting might reduce the number of reports, our own independent examination of SARs suggested that it would be likely to strengthen the quality of reports. Raising the threshold in this way may have the effect of a limited reduction on the volume of disclosures, but should also lead to a wider enhancement of the quality of the intelligence and amount of raw data provided for every SAR submitted. By amending the threshold, reporters would be directed to substantiate their suspicion on objective grounds. This would address the concerns raised by the SFO, for example, that there are too many authorised disclosures which lack real evidence and detail of the suspicion.
5.112 Whilst the concern about reducing the volume of intelligence is clear, it draws attention to a broader issue; the absence of any large-scale and comprehensive analysis of both required and authorised disclosures on which to base those conclusions. Those who responded were unable to provide evidence for their assumptions.
5.113 As we previously stated, during our data analysis work on authorised disclosures, 15% of disclosures lodged failed to meet the existing legal threshold of suspicion and 47.6% failed to provide one or more objective grounds in support of their suspicion. As such, statements from consultees suggesting that there is little to no room for improvement are undermined by the evidence that we have presented above.
5.114 Our analysis of the quality of reports demonstrated that some SARs are being lodged at present with an objective basis for the suspicion and some are lodged without any such grounding. In some cases, there may be objective grounds on which to base the suspicion but the reporter has failed to articulate them. In these circumstances, law enforcement agencies are not able to benefit from potentially useful intelligence. In other cases, there may be no objective basis for the suspicion. In some of the reports that we saw, irrelevant and, in some cases, discriminatory factors such as the nationality of the account holder were the basis for the report.
5.115 Where there are already objective grounds for a suspicion but they have not been articulated, it is unquestionably better for law enforcement agencies to have access to this information without having to make further enquiries. This saves time and scarce resources. Where there are no objective grounds for a suspicion, the value is more difficult to assess. A hunch may, in time, prove to be correct. Likewise, it may corroborate or relate to another reporter's suspicion and together both reports may provide a fuller picture of criminality. However, the lodging of SARs where they are based on irrelevant or discriminatory factors cannot, in our view be justified, even if law enforcement agencies would prefer to have such information. The consequences for the subject of the disclosure are potentially severe. Neither would this be proportionate.
5.116 However, our own small-scale data analysis and the responses of some consultees has highlighted that further research in this area would provide definitive evidence on which to make any decision on amending the threshold. In view of law enforcement agencies' concerns that this change may «turn off the intelligence tap», we are recommending a staged approach based on a continuing commitment to measure the volume and value of SARs. Our recommendations above - statutory guidance on suspicion coupled with a prescribed form - will target those authorised disclosures which do not currently meet the Da Silva test. We also recommend that further empirical research is undertaken on the quality of required and authorised disclosures involving law enforcement agencies. The Advisory Board that we have recommended could then consider, in time, whether to pilot a new threshold test in order to provide a sufficient evidence base on which to build a stronger and more proportionate regime without sacrificing vital intelligence and raw data.
Recommendation 9.
5.117 We recommend that an Advisory Board should undertake a review as to whether to increase the threshold after further empirical research on the quality of required and authorised disclosures is completed.