Publication - Report

Effectiveness of actions to reduce harm from nuisance calls in Scotland

Published: 19 Mar 2018
Directorate:
Economic Development Directorate
Part of:
Economy
ISBN:
9781788516532

Research commissioned to analyse the impact of actions set out in the Nuisance Calls Commission action plan, and to examine the outcomes of past interventions.

144 page PDF

2.3 MB

144 page PDF

2.3 MB

Contents
Effectiveness of actions to reduce harm from nuisance calls in Scotland
4 Future monitoring of effectiveness of actions

144 page PDF

2.3 MB

4 Future monitoring of effectiveness of actions

The previous chapter looked at the likely effectiveness of actions in the Scottish Nuisance Calls Action Plan. The Scottish Government has also requested a review of relevant measurements that could help it [87] to monitor the actual effectiveness of the Action Plan over the next few years.

Earlier chapters and associated annexes have already discussed existing measurements that we are aware of (see, in particular, Figure 2 and Annex H). This chapter aims to bring together this material, with some additions, so as to highlight gaps which could usefully be filled.

Figure 25 summarises the main measurements that SG could make or request, which seem likely to help with assessing the effectiveness of different elements of the Scottish Action Plan. Several will need to be complemented by external data or estimates. For example, it is relatively easy to measure whether Scottish Government schemes are complying with best practice for not stimulating nuisance calls (once best practice has been codified); but what effect this has on the level of nuisance calls in Scotland will be a matter for informed judgement. Progress should ultimately be reflected in lower levels of nuisance calls and of harm.

Relevant conclusions and recommendations are included in the final chapter.

Figure 25: Approaches to monitoring effectiveness of Scottish Action Plan

Action in the Scottish Action Plan Approaches to measuring and monitoring effectiveness
A Consumer protection and empowerment
A1 Provide call blockers to vulnerable consumers Reports from involved Local Authorities, drawing on end user feedback and independent project evaluation.
A2 Raise awareness of protection options TPS registration data (by Local Authority), CAS feedback, consumer surveys.
A3 Measure impact of this action plan Everything in this column contributes.
B Consumers’ own initiatives
B1 Sign up to TPS TPS registration data, preferably with refreshed estimates of TPS effectiveness.
B2 Block unwanted calls Take-up reports from blocking providers (networks, devices and apps); consumer surveys.
B3 Check before you tick Consumer surveys.
B4 Complain Complaints statistics, consumer surveys, enforcement feedback.
C Business behaviour
C1 Raise awareness of the rules Annual report to SG from lead business body on relevant activities and outcomes, including display of number for outbound calls (see D3).
C2 Build partnerships with financial providers to protect vulnerable consumers
C3 Encourage best telemarketing practice
C4 Include vulnerability in the Business Pledge
D Government response
D1 Update SG impact assessments to include consumer impact Effects too remote to be measurable.
D2 Ensure SG schemes meet best practice in not stimulating nuisance calls SG publicity department to check and report schemes’ compliance with best practice guidelines.
D3 Display a number for SG outbound calls SG telecoms department to implement and report.
D4 Work to improve regulatory solutions:
  • Consider making live voice calls illegal unless opted-in
ICO to report on the situation under GDPR.
  • Get telcos to block more where technically feasible
See B2.
  • Make complaining easier
See B4.
D5 Develop a scams prevention strategy Include and identify phone scams, so that evaluation of the strategy will show developments on phone scams.

4.1 Measuring the underlying level of nuisance calls

As was stressed early in this report, it is vital to measure the underlying level of nuisance calls if we are to get any idea of the effectiveness of counter measures. Variability in this underlying level is probably large enough to mask the effects of actions to reduce the level. Annex C attempted to estimate the level from published information, and got no consistent results.

Until recently, survey data on nuisance calls received have been used as a proxy for the underlying level of nuisance calls. With increasingly effective call suppression, this approach is no longer valid; and fortunately Ofcom, with the Nuisance Calls MoU Group of network operators, are well placed to produce and share estimates of this underlying level, using their monthly measurements and other observations available to network operators.

Information exchanged within the MoU Group has been regarded as commercially confidential, even in aggregate (so that no one operator’s information is identifiable). We believe that everyone concerned with reducing harm from nuisance calls, not least the operators themselves, would benefit from having base measurements against which progress could be assessed. We therefore recommend that this group should jointly come up with periodic evidence-based estimates of levels of nuisance calling that could be shared with all concerned.

Considerable value could be added by estimates of various breakdowns of the totals, in particular:

  • Geographic breakdowns: for example, the Scottish Government wants to focus on nuisance calls targeting people in Scotland. Geographic information is embedded in the call details already being collected, although more processing would be needed to extract it.
  • Call characteristics: these may not be evident simply from the collected call details, but by sharing the findings of related investigations, operators could doubtless put together reasonable ideas of whether calls are silent or abandoned, and live or recorded. Often, it should also be possible to establish the likely origin and purpose of the calls.

Call centre experts like ContactBabel are also well placed to estimate levels of calling to and from UK-based call centres (and associated overseas operations), based on industry surveys and detailed sectoral knowledge. ContactBabel produce and sell an annual report which includes such information. The Scottish Government may want to explore options for accessing such information regularly.

4.2 Measuring nuisance calls received

We have discussed at length two main existing sources of information on nuisance calls received:

  • Consumer surveys (to date mainly carried out by Ofcom), in which sampled consumers either recall or record their experience of nuisance calls. Which?, Uswitch and BT from time to time have also carried out and published the findings of similar surveys.
  • User equipment or apps designed to help users control the calls they receive. These may assemble aggregate data on nuisance calls reaching consumer connections and their fate when they arrive, for example whether they are suppressed or answered. The main source of UK data for this study has been the call blocker company trueCall; some (international) data have also been available from app companies Truecaller, hiya and First Orion.

The effectiveness of mechanisms for suppressing nuisance calls, whether in networks or in user equipment or apps, can be assessed in two ways:

  • By comparing estimated base levels of nuisance calling with consumers’ reported experience of receiving nuisance calls. The difference should represent calls which are suppressed before they reach consumers.
  • From companies’ own reports of their achievements in this area. To date these reports have been highly selective and occasional, reflecting companies’ commercial motivations.

These two approaches combined should give better results than either approach alone. Synchronising surveys and company reports (so that they both refer to the same week or month) would be helpful.

A company’s success in nuisance call suppression may well become a factor for some consumers in choosing between companies. However, consumers cannot be expected to assess all the features of these technologies, so it is highly desirable that the companies’ reports on their achievements should be in comparable form, and also that they should be independently verifiable.

Once network-based estimates of the underlying level of nuisance calls become available, the surveys can be freed to look at user experience, in particular what incidence and content of nuisance calls annoys or worries people. Distinctions by age, gender, working status and socio-economic group are helpful. The landline nuisance calls diary survey could usefully be complemented with an analogous survey on mobile nuisance calls and texts. The questions in the consumer issues survey that ask people to look back over the previous four weeks seem of relatively little value, as they may well be answered wrongly.

The user device and app measurements have continuing value for both landline and mobile services, because by crowd sourcing they could help to identify new forms of suspicious or fraudulent call.

4.2.1 Consumer awareness and take-up of protection options

As has already been stressed, the effectiveness of opt-in call suppression depends crucially on consumers’ own decisions to opt in to them, which in turn depends on their being aware of these options. Service, app and device providers may themselves measure consumer awareness as well as take-up, and SG may gain their co-operation in sharing such measurements.

Citizens Advice Scotland and Which? should be able to report on the reach of their September 2017 awareness campaign. In future they might increase the interactivity of their online advice pages so as to be able also to report on which pieces of advice consumers view the most, and find the most useful.

Ofcom’s March 2015 consumer research into awareness and take-up of protection options (see Annex E) has been valuable for this study. Ofcom is currently reviewing its market research on nuisance calls [88] , and one outcome of the review might be repeating this set of awareness questions at intervals. As well or instead, SG could itself commission consumer research in this area.

4.2.1.1 TPS registration data

In August 2017, TPS provided to this study their data on how many telephone numbers in each geographic area code in Scotland are on the register. These data could be provided at regular intervals, and translated into “ TPS take-up” figures for each local authority area in Scotland. Figure 26 shows our initial translation of this kind; Annex L explains how it was done. The average take-up for Scotland as a whole is 73 per 100 households, and the figure shows that TPS registrations per 100 households are mostly between 65 and 85, but there are deviations in each direction, which may help in focusing local awareness campaigns, either for individual council areas or for particular intermediate zones [89] .

Regular repetition of this exercise could shed light on the effectiveness of consumer awareness campaigns in different local authority areas, even if difficulties in attributing the TPS figures to local authority areas cast doubt on some individual area results [90] .

Figure 26: TPS registrations per household by council area in Scotland

Figure 26: TPS registrations per household by council area in Scotland

4.3 Measuring consumer harm resulting from nuisance calls

4.3.1 Complaints and case data

A traditional indicator of consumer harm (for nuisance calls and in many sectors) is complaints. Because they reflect harm experienced, complaints about nuisance calls have the potential to be an especially valuable measure for the Scottish Government. The extra data that complainants are asked for can also be a useful source of intelligence for enforcement purposes.

Figure 27 shows how complaints to ICO [91] , Ofcom and TPS have varied since 2010. Following a large peak in 2013 (when the issue received much publicity and ICO made it easier to complain), there have been apparently random variations and little discernible overall trend. Figure 1 however, complied on an annual basis, shows reductions between 2016 and the first half of 2017.

It is well known that complaints represent only the small tip of a large iceberg [92] , and that complaint levels are affected not only by harm experienced but also by factors such as publicity around the issue in question, people’s awareness of how to complain and the ease or difficulty of making a complaint. Still, as they are collected across sectors and causes for complaint, they are often used to compare consumer experiences across sectors and types of problem.

Figure 27: Monthly complaints to official bodies, 2010 to 2016

Figure 27: Monthly complaints to official bodies, 2010 to 2016

Annex I provides information on the different complaints systems currently in use. As the claims management regulation review noted [93] :

“The existing environment for reporting complaints about direct marketing is confusing for consumers; different types of complaints currently need to be made to one or more of ICO, Ofcom, TPS, CMRU and/or the Advertising Standards Agency ( ASA). The creation of a central reporting point for such complaints could help to reduce such confusion.”

Complaints about spam texts can be made relatively easily, by sending a message from the mobile phone affected to ‘7726’. Figure 28, based on reports of completed cases, shows that this easier reporting makes a big difference - complaints about texts are much more likely to be made by sending a message to ‘7726’ than by contacting ICO, and, overall, taken-up complaints about texts are more than six times as common as taken-up complaints about calls [94] . Yet without the ‘7726’ option, the reverse would hold - at around 4 per 100,000, taken-up complaints per text made directly to ICO are only half as common as taken-up complaints per call.

Figure 28: Modes of complaining to ICO from completed case reports, 2014-2017

Medium Number of complaints cited in completed case reports Proportion of complaints made directly to ICO Number of calls or texts causing complaints Complaints per 100,000 calls or texts causing complaints
Calls 1,294 78.13% 16,707,773 7.7
Texts 10,361 7.71% 20,377,862 50.8

To provide another comparison, BT has said that since the launch of BT Call Protect, there are 80,000 calls per week (or at least 320,000 calls per month) to ‘1572’ to add numbers to personal blacklists, change settings or check the messages left by calls that might be nuisance calls [95] . Despite arising from just 2 million service users, this number dwarfs the average number of complaints made to ICO, Ofcom and TPS together, which was about 6,200 per week in 2016 [96] .

We recommend that the two organisations mainly receiving complaints about nuisance calls, ICO (now in overall charge of TPS) and Ofcom, jointly review how to maximise public value from their complaints statistics. Changes for consideration include:

  • Finding the most meaningful period over which to aggregate complaints, which is long enough to eliminate “noise” but short enough to reflect genuine trends.
  • Synchronising complaints publication from the two organisations, using consistent breakdowns, and combining figures when this makes sense so as to provide a single overall indicator.
  • Providing their best interpretation of the factors affecting changes in complaints levels, such as changes in nuisance call levels or types, availability and take-up of call suppression services, effective enforcement action, relevant publicity, or changed complaints systems.

Action Fraud and Police Scotland, working with Trading Standards Scotland, also gather complaints and case data on scam calls. These data should also feed in to the Scottish Government assessment of its Action Plan, and as far as possible be integrated with the ICO/Ofcom exercise just outlined.

The Scottish Government may wish to request the UK organisations to pass on data from the subset of complaints that can be identified as from Scottish complainants or about Scottish originators. If complaints from Scottish complainants were to fall faster than complaints from the rest of the UK, it would be a valuable indicator of success of the Action Plan.

4.3.2 Consumer feelings about nuisance calls

Ofcom’s diary survey, and occasionally other surveys, ask people how they find the nuisance calls that they receive (choosing among distressing, annoying, not a problem and useful – see Figure 29). Since consistently over 80% of respondents find the calls annoying, changed levels of distress may be hard to detect with any confidence. As is shown in Figure 30, distress is largely (though not only) attributable to scam calls, and it may work better to explore this as part of the Scottish Government’s broader scams prevention strategy than as a subset of nuisance call research. (Note that the kinds of calls shown in the middle part of Figure 30 were chosen by respondents; many of those shown as financial services, computer maintenance etc may in fact also have been scam calls).

Figure 29: Feelings about nuisance calls overall, 2013 to 2017

Figure 29: Feelings about nuisance calls overall, 2013 to 2017

Source: Presentation slides of Ofcom landline nuisance calls diary survey 2017

Figure 30: Distress caused by nuisance calls

Percentage of 2017 diarists finding calls of certain kinds distressing
Abandoned calls 13%
Silent calls 9%
Recorded sales calls 8%
Live sales calls 7%
Scam calls 24%
Financial services 16%
Computer maintenance 11%
Accident claims 10%
PPI 8%
Home improvements 7%
Top reasons given by 2017 diarists for finding calls distressing
They keep phoning 20%
Silent calls / caller hangs up 20%
Disturbed / time wasting 12%
Scam call 11%

Source: Ofcom landline nuisance calls diary survey 2017

Other sources for assessing how consumers feel about the calls they receive include MPs’ postbags (in recent years nuisance calls have been a major element), and qualitative reports from call suppression providers. For example, BT and TalkTalk, who have been very active in phone scam prevention, may be able to report on the volume and nature of calls they get to their advice lines; while trueCall and others may provide verbatim comments from their customers who have blocked certain callers, as well as reflecting to SG any value obtained from public crowdsourced databases like whocallsme.com.


Contact