Justice CommitteeWritten evidence from Dr Francis Beresford

This submission refers to three documents that give supporting information:

Doc 1.Review of FWA and alternatives.3.comb.doc

Doc 2.To Sir Alan Beith in response to Peter Handcock.2.8.12.doc

Doc 3.Reply to Iain Bell.Chief Statistician.18.7.12.doc

Summary

1. Since the Runciman Commission made its recommendations in 1993, trained and qualified interpreters who are on the National Register of Public Service Interpreters (NRPSI) have provided a high quality interpreting service to courts and police in England and Wales in a system which has gained renown though out Europe.

2. Though interpreting costs escalated from 2004 when the A8 countries joined the EU they have been roughly stable since 2008.

3. In 2010 the Ministry of Justice (MoJ) Interpreting Project identified three concerns with the current arrangements: time consuming and costly booking arrangements; limited availability in some languages; and some areas and complaints not being investigated thoroughly.

4. The MoJ decided to outsource interpreting to a single interpreting agency under a Framework Agreement (FWA) rather than employ interpreters direct to try and reduce costs and improve booking efficiency and complaints.

5. To try and increase the pool of interpreters and reduce costs the MoJ allowed Applied Language Solutions (ALS) to introduce a tiered system where interpreters with language qualifications but no training in legal terminology could interpret for courts and police as Tier 2 interpreters. A superficial assessment was introduced to help this grading.

6. No independent assessment was commissioned to check whether those fulfilling the new criteria could perform adequately in situ even though this seemed unlikely given that they might have no training in interpreting legal terminology.

7. The FWA and introduction of Tier 2 interpreters has in fact produced a dramatic decrease in the available pool of interpreters from the 2,350 interpreters previously on the NRPSI list to the 1,500 currently working for ALS after a year of recruiting.

8. The reduction in training and skills of the interpreters working in courts has been even more dramatic with only 13% of those previously on the NRPSI list now signed up with ALS.

9. The MoJ’s assertion that since many of ALS’s interpreters speak more than one language this increases the effective number of interpreters to 3,000 (ie a 100% increase) only lays bare the marked drop in standards, as the equivalent increase in the NRPSI list when adding multiple languages is only 18% due to the high level of interpreting skill required.

10. The much quoted assertion by Mr Crispin Blunt that interpreters were previously overpaid and could earn £100,000 under the National Agreement is based on a single unsubstantiated quote in the Birmingham Mail repeated in the Sunday Times and has no basis in fact.

11. Average interpreter pay in England, prior to the FWA, was £15,000 with a small percentage of interpreters working in a few specific languages, earning over £30,000.

12. ALS has reduced the average hourly rate for a freelance interpreter, including travel time, from £23/hr under the previous National Agreement to an average £11/hr under the FWA.

13. The evidence that outsourcing might reduce costs came from the Collaborative Police contract in the North-West but this initially offered higher rates of £30–35/hr to trained interpreters, was on ALS’s home territory where it had the necessary contacts and was only for police work which is not observed so poor standards could go unrecognised.

14. Even though it was inviting bids for a £60 million pound a year contract the MoJ had no national data on the number of interpreting assignments it required or in which languages or what locations.

15. As the MoJ nonetheless asked for bids for an hourly rate to include travel time and expenses it was always likely that bids would be based mainly on guesswork and would end up too high or too low.

16. ALS’s tender submission stated that “approximately 2,500 of our 4,500 registered freelance are suitably experienced and qualified for Authority assignments” when in fact most of those 2,500 were on an NRPSI database which ALS had bought. Only a tiny percentage of these interpreters had actually agreed to work for them.

17. ALS and the MoJ could only claim they were paying interpreters “market rates” which is usually defined as “the usual rate in the market” as ALS’s Interpreter Working Group were deciding themselves what the figure should be.

18. The MoJ appears to have simply taken without question the things ALS was promising in the contract and not obtained suitable independent expert advice. For instance the first Key Performance Indicator (KPI) is “availability of all languages within a 25 mile radius—95%”. This is an impossible target as many languages have less than ten NRPSI interpreters for the whole country.

19. ALS was awarded a £60 million/year contract by the MoJ even though in May 2010 it had a turnover of £7.5 million and losses of £20,000 and in May 2011 a turnover of 10.6 million and losses of £331,000.

20. The first month of roll out in Feb 2012 saw interpreting chaos in the courts and those police forces who had signed up with over one third of interpreting assignments not being fulfilled.

21. Though overall rates fulfilment rates have increased this has almost certainly only been possible only as ALS has been ignoring the quality requirements for Tier 2 interpreters.

22. The MoJ has not published data on fulfilment for the less common languages, which are where ALS is likely to have major long term problems in fulfilment, despite requests, presumably as this would should how severe the shortage of interpreters still is.

23. It is likely that most of ALS’s Tier 2 interpreters, other than those with partial DPSI, do not fulfil all the requirements stated in the contract ie did not have 100 hours interpreting experience before starting work, plus a language related degree, plus previous experience of working in a justice system and should not therefore be working.

24. It is also highly likely that a substantial number of Tier 2 interpreters do not have the required enhanced CRB check within three years or the equivalent and some may have no CRB check at all which would also invalidate them from working.

25. Of the 720 applications ALS made between Jan and July 2012 for Non Police Personnel Vetting (NPPV3) a remarkable 20% were turned down which speaks volumes about the calibre of ALS’s workforce.

26. The contract states that interpreters/translators will be covered by professional indemnity but in mid-April 2012 the ALS website argued that this was not necessary, though this was subsequently removed. It is not yet clear if ALS has complied with this part of the contract.

27. Reports of interpreters not being informed of the result of their assessments suggest that ALS may not even have been arranging for all of these to be marked by Middlesex University.

28. The MoJ has up until now steadfastly refused to admit there are any problems of interpreter quality relating to the FWA. However in early August Mr Alan Handcock, Chief Exec of the Courts and Tribunals service was informed personally and in detail both of the problems and the questions that need to be put to ALS to confirm if it is complying with the quality aspects of the contract as above. Hopefully these will now be addressed.

29. The huge reduction in the standard of interpreting detailed above goes a long way to explain why so many of those working in the CJS and from all sections of the press are critical of the FWA. It is of note that all the senior management in ALS who were involved in negotiating the FWA and introducing the Tiering system which has proved so disastrous for interpreting standards in this country have now left ALS/Capita and cannot get further jobs in the language industry in this country. Why is the MoJ still clinging to this failed experiment, particularly when other options which save money are available? (Doc 1, p 10&11).

1. The Rationale for Changing Arrangements for the Provision of Interpreter Services

1.1 In 1993 the Runciman Commission recommended that only trained and qualified interpreters be used in court and subsequently the National Register of Public Service Interpreters (NRPSI) was established (1994) along with the Diploma in Public Service Interpreting (DPSI) with legal, community and medical options. In 1998 the Trials Issues Group recommended only NRPSI interpreters be used for criminal investigations and court proceedings, and in 2001 the Auld Report recommended a review of levels of pay “to encourage more and the best qualified to undertake this work”. The high standards of interpreting in England and Wales, and in particular the NRPSI and DPSI, have previously been held up as beacon of good practice to other countries in the European Union by EULITA, the European Union Legal Interpreters and Translators Association.

1.2 Interpreting costs escalated in 2004 when the A8 countries joined the EU but the Ministry of Justice’s published interpreter costs for Crown Courts showed only a marginal increase between 2008–2010 from £4.5 million to £4.6 million. In 2010 the MoJ established the Interpreting Project to improve the existing system and reduce costs while maintaining appropriate quality standards. The review (381–11 03 30 Document 5—Draft Quality Impact Assessment -30 March 2011 p3) identified three major concerns: (a)the time consuming and costly mechanism for booking interpreters (b)the limited availability of registered interpreters in some languages and in some parts of the country (c) that complaints made against interpreters were not being investigated thoroughly.

1.3 (a)The booking system could be time consuming and inefficient as there was no local or national coordination to enable bookers to know which interpreters were available (other than partly for Tribunals) and fees for travel time, travel expenses were negotiated individually for each job (other than for Tribunals). (b)The perception of limited availability of interpreters in some languages and some areas was based on heresay, as no audit was undertaken, and was at least partly explained by the extremely poor search system on the NRPSI website which meant interpreters in neighbouring counties, even if close by, did not appear on the searches. The NRPSI board was also charging a substantial fee for access to the register which meant not all parts of the CJS had access to it. (c) The NRPSI board which was responsible for investigating complaints had neglected this role for some time and had a considerable back log as it was consumed by legal battles with interpreters. In 2011 the Institute of Linguists (IoL) reformed the management of the NRPSI which is now independent of the IoL. The register is now freely available to anyone and has reasonable search facilities.

1.4 The Interpreting Project hoped to reduce travel expenses for interpreters by widening the pool of interpreters beyond those on the NRPSI, and thus hopefully increasing the distribution of interpreters. It was also persuaded by ALS that less highly trained interpreters ie without the DPSI could perform adequately in some court and police cases, although there was no actual evidence for this, and would also accept lower rates of pay. Thus was born the concept of Tiers of interpreters where Tier 1 interpreters had the DPSI or Met test and Tier 2 interpreters had either past the DPSI exam apart from the written test, or had a language related degree with an English component or relevant diploma plus previous experience of criminal justice work plus at least 100 hours of public service interpreting experience. Tier 3 interpreters were meant to be only used for community work. A new assessment was introduced to support this Tiering which mimicked the DPSI in a very superficial way but was machine based and included no in-depth assessment of legal terminology or interpreting skills (Doc 2, p.6 ).

1.5 Having decided to re-write the requirements for court and police interpreters contrary to all the previous recommendations the MoJ did not then undertake a trial to see if interpreters meeting these new standards could perform adequately in courts or police stations. It also did not set any limits on the situations in which Tier 2 interpreters were used but left this up to individual justice organisations to decide. As it happened, as very few Tier 1 interpreters signed up with ALS, apart from very high profile cases the courts and police will have been forced to usually mostly Tier 2 interpreters.

1.6 Instead of increasing the pool of trained interpreters as envisaged the FWA has in fact managed to dramatically reduce both the pool of interpreters available and their skill base. The pool of 2,350 NRPSI interpreters available prior to the FWA, has shrunk to 1,500 interpreters of whom only 301 have the DPSI (in July 2012). The MoJ has tried to put a positive spin on this by saying that as some interpreters speak more than one language this equates to 3,000 interpreters by language (ie an increase of 100%). However this simply shows how dramatically standards have fallen as very few people manage to master a second language sufficiently to pass the DPSI, the commonest combinations being Urdu/Punjabi and Mandarin/Cantonese. When including multiple languages the number of DPSI interpreters only increases by 17%. Interestingly, and also contrary to the original aims, as the rates of pay are now so low and there is no three hour minimum many ALS interpreters will now not take court jobs near to where to they live as jobs with more travel time give a more a guaranteed income.

1.7 Mr Crispin Blunt has promulgated a myth, both in the House and on the radio, that interpreter pay was excessive under the National Agreement and that interpreters earned more than £100,000. In fact under the National Agreement, with a three hour minimum, the average total hourly rate, including travel time was £23/hr and for the Tribunals Service was around £20/hr (Doc 1, p 5). According to a survey by the Institute of Translation and Interpreting (ITI) in 2011, average annual income for interpreters in this country was £15,000. Interpreters working in languages such as Slovak & Romanian, where the demand is usually more than the supply, can earn around £30,000—£35,000 from the National Agreement. A very small number of interpreters earn more than this by working very long hours. The assertion that interpreters earned more than £100,000 appears to be based on an article quoting a single anonymous interpreter in a pro ALS article in the Birmingham Mail (23 September 2011) which was repeated in the Sunday Times (23 October 2011). From the hourly rates above, even given overtime, it is clear this is virtually impossible, and no authentic case has yet been found. The ITI survey does identify a small number of interpreters/translators earning over £100,000 but these either run their own interpreting companies or were doing work outside the National Agreement which pays much higher rates.

2. The Nature and Appropriateness of the Procurement Process

2.1 The original aim of the review of interpreting and translation services was to make savings of 10% on an approximate £60 million spend. (e-mail by Richard Mason to Interpreters 30 March 2011). This was a reasonable target but was subsequently forgotten following Applied Language Solutions’ (ALS) promises of savings of £18 million.

2.2 One can see that at the time a Framework Agreement with a single provider seemed like a good solution for the MOJ. Though negotiation with interpreters organisations might have produced the 10% reduction in costs that was being aimed for it would not have dealt with the problems associated with booking interpreters. Expanding the Tribunals Service Booking Service, which was well run and highly cost-effective, to cover the whole of the CJS was probably too “hands on” a solution to be politically acceptable. A solution that uses a nationwide web based booking system (Doc 1, p 10) has recently been proposed which would solve the problem but this idea had not been put forward in 2011. A Framework Agreement delegated the responsibility for the organisation of interpreters, negotiating reduced terms and ensuring adequate standards to a private company and out of the hands of the MOJ. It also appeared to offer a solution to the difficulties with bookings and complaints and ALS also promised continuous professional development and appraisals. A single contractor gave the selling point of a single contact phone number, though for a service not used by the public this would seem to have little advantage. Ironically in its tender ALS charges extra for phone bookings to discourage them.

2.3 There were however potential problems and downsides to this decision as interpreters pointed out in their responses:

All the large agencies in England had in reality up till then done most of their interpreting work in one or two regions in the country. This meant national roll out for a single contractor was always going to be difficult as local contacts are vital for interpreting agencies.

The MOJ, as a virtual monopoly employer of interpreters used in a legal setting (over 90% at a rough estimate), has a strong motivation to develop a sustainable, high quality service interpreting service. In contrast an agency on a fixed term contract is naturally going to look at maximising its short term profits and to be less concerned with issues of long term standards or provision. This is particularly true for a single provider where there is no competition to drive up standards.

Though a single provider would administratively be much more cost efficient at booking and paying interpreters the MOJ was paying by assignment not in bulk. In order to maintain accounting controls the courts and police would therefore still have considerable administrative costs recording jobs done and checking the agency’s invoices against these.

The Government had recently stopped employing freelance consultants via agencies as it added a considerable extra layer of expense. It was equally probable that this extra layer of expense would apply if an agency employed freelance interpreters and lead to reduced income for interpreters.

2.4 The main evidence that a Framework Agreement might succeed came from the North West Collaborative contract for interpreting which significantly reduced headline interpreter costs for Greater Manchester Police and other North–West forces. This was however achieved by ALS paying some NRPSI interpreters £30–35/hour for interpreting time without travel time or travel costs (though its initial proposed rate was lower) and non NRPSI interpreters lower rates. Though the contract had a difficult first two months it succeeded in attracting just enough NRPSI interpreters, supplemented by non-NRPSI interpreters already known to ALS, to fulfil the contract. There were however some particular features of this contract which meant that a similar contract, particularly one offering considerably lower rates, was likely to be much more difficult to roll out either to courts or nationwide.

(1)It was on ALS’s “home territory” where ALS already had many non-NRPSI interpreter contacts.

(2)It was only for police jobs which, in 90% of cases do not last more than 2 hours, and can wait a while for an interpreter, unlike courts. This meant ALS could sequence jobs relatively easily and so get away with a relatively small pool of interpreters.

(3)If no travel time is paid and minimum time is one hour, two hour police jobs give a significantly higher hourly rate of pay to interpreters than a one hour magistrate jobs.

(4)Police interviews usually have no witnesses and the North-West police were either not encouraged or actively discouraged from making complaints about any reductions in the standard of interpreting. Therefore though many examples of non-fulfilment and poor interpreting from use of non-NRPSI interpreters were collected by NRPSI interpreters these did not reach the press. The openness of courts to the public and the presence of judges meant courts were unlikely to be so tolerant of reduced standards.

2.5 Though the MoJ made the decision to appoint a single national provider it unfortunately had minimal information for potential bidders on either the total number of assignments required, or their geographical and language distribution. The companies however seem to have been asked to quote for an hourly rate for each Tier of interpreter inclusive of travel time and travel expenses. Though this payment method was obviously convenient for the MOJ in calculating fees per case, interpreters had traditionally been paid travel time and travel expenses and these were vital to make jobs involving minority languages, where travel is more likely, and geographically isolated sites, viable. Travel time and expenses were also needed to allow approximate equalisation of hourly rates of pay for assignments of different duration and location. As the agencies had no information on these costs their bids were always going to involve a large amount of guesswork. It is reported that ALS, a relative newcomer to the interpreting industry (2003), bid substantially lower than the other two more experienced agencies.

2.6 By March 30th 2011 when interpreters were consulted on the new Framework Agreement it was clear that the MOJ was having considerable problems realising which of the improvements it was asking for were practically achievable and which were not and it was also clear that ALS was not choosing to enlighten them. For instant the first key performance indicator in the proposed contract had as its objective the absolutely impossible target of “availability of all languages within a 25 mile radius—95%” (over 140 languages are needed and many have less than 10 NRPSI interpreters to cover the country!). The MOJ also appears to have been taking little account of interpreters’ responses to the proposed FWA as despite the impossibility of this target being pointed out in interpreters’ responses it still made its way into the final contract.

2.7 The tender submission by ALS included highly misleading statements and also promises it has subsequently made no attempt to honour or has ignored from the start of the contract. For example in its Tender Response ALS states that “approximately 2,500 of our 4,500 registered freelance interpreters are suitably experienced and qualified for Authority assignments” (page 114). What ALS fails to mention is that this number is almost entirely made up of the database of NRPSI interpreters which it had purchased. ALS would have known that most of those interpreters had neither any idea they were on their database nor had agreed to work for them. The MoJ subsequently became painfully aware of this lack of available interpreters in February 2012. ALS’s ignoring of standards for Tier 2 interpreters, CRB checks and indemnity insurance for interpreters will be dealt with later.

2.8 Both ALS and the MOJ have repeatedly asserted that under the Framework Agreement they are paying interpreters “market rates”. In its tender response ALS states that “one of its unique selling points” is “before and throughout the economic crisis we have continued to pay our linguists above the market rate for their services.” The official definition of “market rate” is “the usual rate in the market”. However when using the term in the Tender Response (Framework Agreement page 126) the phrase does not have its usual meaning but is defined as the rate that ALS’s Interpreter Working Group, a group of interpreters employed by ALS, decide it should be. For instance a private request in April 2011 for one hour of a Polish NRPSI interpreter’s time, gave quotes from Prestige Network and the Big Word, two of the largest agencies in the North-West, of £150 and £120+mileage respectively, whilst ALS quoted £45 + £20/hr travel time + 40p/mile (all excl. VAT). When ALS published its proposed rates for interpreters under the FWA in Sept 2011 these gave an average rate of pay of £11/hr including travel time (Doc 1,p 5/6) which is less than half the rate under the National Agreement. This clearly shows the extraordinarily large drop in pay ALS was demanding of trained interpreters and its disregard for previous market rates and easily explains why 87% of NRPSI interpreters have not signed up with them.

2.9 It also remains a puzzle how a contract worth around £60 million a year to a company which in May 2010 had a turnover of £7.5 million and overall losses of £20,000 and in May 2011 a turnover of £10.6 million with overall losses of £331,000 passed the MoJ’s pre-contract financial scrutiny checks.

3. The Experience of Courts and Prisons in Receiving Interpretation Services That Meet Their Needs

3.1 By the start of 2012 over 1,000 NRPSI interpreters had signed a petition refusing to work for ALS and the start date of the contract, originally meant to be in September had been repeatedly delayed. Though this should have alerted the MoJ to the problems that ALS were having recruiting enough interpreters they pressed on regardless and as expected the national roll-out in February 2012 caused chaos in the courts with very large numbers of cases unfulfilled.

3.2 Though the overall fulfilment rate has improved since then from 65% in Feb to 90% in April there is a significant possibility ALS has manipulated these figures by misallocating some “unfulfilled cases” to “cancelled by customer” (Doc 3,p 5). This can easily be proved or disproved by comparing previous Tribunal statistics with the “cancelled by customer” figures in the FWA but the MoJ statistics office have so far not responded to this request (Doc 3, p 7). The MoJ has also only published the success rates for the top 20 languages despite the fact that the majority of the remaining 122 languages are likely to have fulfilment rates well below 90% and in many cases zero, as finding interpreters for minority languages is much less easy. The chief statistician has used a statistical technicality to support non-publication of this data though this technicality is not applied by himself elsewhere in his report or applied by the EU statistics agency or the Northern Ireland Courts and Tribunals service to their data. Publication of this data is clearly in the public interest (Doc 3, p 3).

3.3 The emphasis the MoJ has placed on the improving fulfilment rates under the FWA tends to distract observers from the FWA’s major failing which is that the standard of interpreting is in many cases completely unacceptable for court and police work. As only 301 NRPSI interpreters have signed up with ALS, 1,200 of ALS’s 1,500 interpreters must be Tier 2 or below. The FWA requires Tier 2 interpreters to have either “Partial DPSI in law” (having failed only the written component) or a language related degree or diploma involving English, plus previous work in a criminal justice system whether paid or voluntary, plus a university degree, plus 100 hours of public service interpreting. Though there will be a small number of interpreters with “partial DPSI in law” employed by ALS who fulfil these requirements it has been obvious since ALS started recruiting in Sept 2011 that £11/hr for freelance work is simply not enough to attract such educated candidates. Since roll out in Feb 2012 ALS has therefore usually been simply been ignoring these requirements and taking on anyone it can find who has basic language skills and is willing to brazen it out in court when they have little or no knowledge of the legal terminology being used . ALS may have been largely relying, if on any requirement at all, solely on a pass at Tier 2 level in the assessment something the designer of the assessment has specifically stated is quite inadequate. http://www.lawgazette.co.uk/blogs/blogs/news-blogs/interpreting-interpreters-strike#comment-14643Some carefully phrased questions to ALS will quickly establish the facts in this matter (Doc 2, p 4/5) and it is likely that very few of the Tier 2 interpreters without partial DPSI in law will fulfil the stated requirements. This would make a sizeable percentage of ALS’s workforce ineligible to interpret in court or the police. There is also considerable doubt, even if all the Tier 2 requirements were fulfilled, whether many Tier 2 interpreters would perform adequately in court as the requirements do not ensure detailed knowledge of legal terminology.

3.4 The other major area in which ALS has been failing to meet the requirements of courts and police is in making sure that all its interpreters have the necessary CRB clearance. The contract states that all interpreters should have an enhanced CRB check as a minimum requirement prior to working. In its Tender Submission (page 150 of the Language Services Framework Contract) ALS states that if an enhanced CRB check is not in place or is more than three years old they will carry out the check. However Interpreters going to the assessment centres in the first month or so described taking their documents and no one at the Centres being interested in them. There has also been at least one email from ALS to its interpreters asking them to upload details of their CRB checks if they have not done so and individual ALS interpreters attending court have also admitted to other interpreters that they have had no CRB check. It therefore seems likely that in its rush to find interpreters to fulfil the contract that ALS as not been complying fully with the requirements for CRB checks. This would again make a proportion of its workforce ineligible to work in court until the necessary checks were made.

3.5 In July 2012 a Freedom of Information Act reply from Warwickshire Police Force, who do CRB checks for the ALS contract, showed that since January 2012 ALS made 720 applications for its interpreters for Non Police Personnel Vetting (NPPV 3) status of whom only 574 (ie 80%) were accepted. This shows a surprisingly high rate of refusal which must reflect on the quality of ALS’s interpreters. Assuming that only the 300 or so NRPSI interpreters and perhaps an extra 200 Tier 2 interpreters are likely to have existing CRB checks this could potentially leave a considerable number of interpreters with either no CRB check or one that is out of date. A few carefully phrased questions to ALS would again elucidate the facts in this matter (Doc 2 p 5).

4. The Nature and Effectiveness of the Complaints Procedure

4.1 Though many people are involved in court cases the complaints procedure allowed only the courts themselves to complain to ALS which meant that solicitors, barristers, probation officers and others using ALS interpreters were not able to complain. Complaints could also only be addressed to ALS themselves and no higher level independent authority was available to deal with complaints that were not answered satisfactorily. Complaints by interpreters to ALS over the last nine months, for instance in relation to ALS’s frequent breaches of the Data Confidentiality act with regard to interpreters’ data have invariably gone unanswered.

4.2 It is also seems possible that ALS has manipulated the statistics concerning complaints about quality in the report “Statistics in the use of language services in Courts and Tribunals Feb—April 2012”. Complaints about “quality of interpreter” for this period were 62 while the complaints about “operational issues” were 422. Operational issues are defined as including amongst other things “incorrect tier assigned (the customer has requested a specific tier of assignment and an incorrectly tiered interpreter was assigned)” ie a complaint about the quality of the interpreting. In practice Tier 3 interpreters should not be sent to court or police stations and the MoJ has not specified that there is any sort of court where a Tier 2 interpreter cannot be used so the situation of an incorrect tier of interpreter being sent should not theoretically exist.

5. The Steps That Have Been Taken to Rectify Under-Performance and the Extent to Which They Have Been Effective.

5.1 The MoJ’s response to the underperformance in the contract has involved allowing courts to book interpreters direct if it was clear ALS could not supply someone, though unsurprisingly only a small minority of NRPSI interpreters have taken up this offer. Other than that the MoJ has sat tight and tried to believe its own mantra that “things were just settling down”. Fulfilment rates have increased, as they were gradually bound to if ALS were allowed to largely ignore the quality requirements for Tier 2 interpreters. The rates for less common languages, which the MoJ has so far refused to publish, are however likely to still be dire.

5.2 The MoJ from Kenneth Clarke down has resolutely refused to consider that there are any issues of poor quality in relation to the FWA and have seemed determined to see the problem in Ken Clarke’s words as “an old-fashioned industrial relations dispute”. So despite being notified of problems by the press, by interpreters, by judges, by barristers, by solicitors and others in the Justice System the MoJ has never asked ALS the precise questions which would uncover the truth about whether the quality standards in the contract are being fulfilled (Doc 2). In early August Peter Handcock, Chief Exec of HM Courts & Tribunals, was notified by email, in detail, of these problems and the questions which when put to ALS would uncover these deficits. Given the forthcoming enquiry he will hopefully now look into the matter.

5.3 The MoJ has also failed to understand that professional interpreters by the nature of their language skills tend to be enterprising and multitalented and that unless the current contract is replaced soon by something more reasonable they will lose most of their current NRPSI interpreters for good. Just as most MPs, if their pay were halved overnight, would leave the job and find other work even if they were very dedicated to politics, unless they saw a chance things would change very soon, so also with interpreters. Should this happen it would condemn the British courts to low standards of interpreting, particularly in minority languages, for years to come. It would also from October 2013, when the Implementation of the European Directive on Rights to Interpreting in Criminal Proceedings is to be implemented, bring down the wrath of the EU, along with sizeable financial penalties, for destroying an exemplary high quality interpreting system and replacing it with one more compatible with the most undeveloped parts of the EU.

6. The Appropriateness of Arrangements for Monitoring the Management of the Contract, Including the Quality and the Cost-Effectiveness of the Service Delivered

6.1 When ALS published its proposed rates for interpreters in September 2011 why did the MoJ not realise that halving the rate of pay for trained interpreters would inevitably be unacceptable to most of them? When over half the work force signed a statement saying they would not work for ALS in Dec 2011, why did the MoJ not quickly produce estimates of cost of an interpreter not turning up so it could estimate when the contract was becoming economically non-viable, if roll out led to significant non-fulfillment?

6.2 Once it became clear in the first few weeks that even for assignments where ALS did provide an interpreter who arrived on time, there were problems with interpreters having to leave for another job before the end of the case, why did the MoJ not take steps to monitor this event via the ALS computer system or by individual reports from courts given that it would not otherwise show up on the system?

6.3 When reports have made it quite clear that in trying to fulfil the contract ALS was often ignoring the quality requirements for Tier 2 interpreters eg a language related degree plus 100 hours of public service interpreting experience plus experience of a justice system why did the MoJ not investigated these complaints itself instead of making out that there was no problem?

6.4 The contract states that all interpreters/translators will be “covered by a current and valid policy of professional indemnity” (item1.11 on page 11) and item 34.2 in the contract implies this will be to £1,000,000. However in mid April 2012 the ALS website stated that “professional liability insurance …doesn’t seem useful at this point …and could make you a more attractive target for a lawsuit” (see www.linguistlounge.org and search on “indemnity insurance”). What proof does the MoJ have that ALS has now taken out insurance for all its translators and interpreters?

6.5 There are many reports of interpreters not receiving the result of their interpreting assessment. What proof does the MoJ have from ALS have that all these assessments have been marked by Middlesex University and that all those interpreting in multiple languages have past the assessment in each language to the required degree? (see Doc 2, p 3&5)

6.6 When every newspaper reported problems the FWA along with judges, barristers, solicitors and almost everyone else in the court system why did the MoJ not stop and ask itself whether all those might be right and ALS’s blarney might be wrong?

6.7 Perhaps the most extraordinary revelation to an outsider, of this whole contract has been the degree to which those who have initiated a new and controversial government contract are also allowed to monitor it without any external body empowered to step in if things are not going right. We all make mistakes. Picking these up quickly and rectifying them is embarrassing but not over costly whilst ignoring all the evidence and doggedly continuing with a wrong decision may end up very costly indeed for the tax payer.

August 2012

Annex A

REVIEW OF THE FRAMEWORK AGREEMENT FOR INTERPRETING & ALTERNATIVE PROPOSALS

Summary

The National Register of Public Service Interpreters (NRPSI) was set up in 1994 along with the Diploma of Public Service Interpreting (DPSI) in Law to improve interpreting standards in the Criminal Justice System (CJS). They have both been praised in as beacons of good practice.

In 2010 the Ministry of Justice (MOJ) established the Interpreting Project to look at reforming the provision of Interpreting Services and to reduce costs. There were concerns about inefficient booking systems, complaints not being adequately dealt with and insufficient numbers of interpreters in some areas. Following the apparent success of the North-West Collaborative Police Contract which reduced police costs in the North-West the MOJ decided to negotiate a Framework Agreement with a single supplier. This included the formation of Tiers of Interpreters so that interpreters without the DPSI exam could be included. A new assessment was introduced to help allocate non DPSI interpreters to Tiers, though its validity was not tested.

The contract was priced by hourly rate per tier and was won by Applied Language Solutions (ALS) who by report bid considerably lower than the other two more experienced agencies. The average overall rate per hour offered to freelance NRPSI interpreters by ALS is less than half the rate they were paid previously (£11 vs. £23). Many NRPSI interpreters have refused to sign up with ALS and there have been considerable problems with non-fulfilment of jobs in courts and police stations. Lower standards of interpreting have also been reported which is unsurprising since most of the newly signed interpreters have had no training in legal interpreting.

In the long term £11/hr is too low a rate to sustain freelance work by postgraduates with higher training. It is likely most of the NRPSI interpreters signed up with ALS will in time find other better paid work. Standards will fall further as there will be no incentive to train for the DPSI as most jobs are Tier 2 and do not require it. Less skilled interpreting is likely to increase costs due to slower speed, fewer early guilty pleas due to poorer communication, and increased retrials and miscarriages of justice. Shortages of interpreters will increase costs due to cancelled cases.

There are doubts about ALS’s probity including dubious representation of interpreter figures, breaches of the Data Protection Act, and doubts about compliance with the conditions for Tier 2 interpreters. The long term financial viability of the contract for ALS/Capita is also unclear.

Other organisations in the CJS have reduced costs considerably without using agencies. If the MOJ were to again employ NRPSI interpreters directly at their previous rates but with a two hour minimum for police work the MOJ would save around 21% compared to previous costs, roughly the same as with the Framework Agreement. Similar savings on minimum time could be made for court cases where the finish time could be reliably estimated. If as a result of these more competitive terms the MOJ directed that all court and police work was done by NRPSI interpreters (as per the Auld Report) there is a good chance interpreters would find these terms acceptable.

Interpreter booking could usually be reduced to a single phone call by having a national booking website on which interpreters indicated their daily availability. By automatically calculating realistic travel time and other fees and automating all paperwork and payments the website would reduce administration costs to less than any agency could achieve. Complaints and postgraduate training could be dealt with via an interpreter manager and the existing professional bodies.

REVIEW OF THE FRAMEWORK AGREEMENT FOR INTERPRETING AND ALTERNATIVE PROPOSALS

15 APRIL 2012

Since its national rollout in February 2012 the Framework Agreement for Interpreting services has been producing difficult headlines for the Ministry of Justice with stories of poor quality interpreting and interpreters failing to appear at courts and police stations. When all factions of the press are united in their criticism of a new system it is likely to indicate there are significant problems. After more than two months of difficulties an in-depth review is required to look at why and how the Framework Agreement was first contracted, why problems have arisen, the outlook for the Agreement and whether there are any alternatives which would still allow the MOJ to make savings.

Background to the Previous National Agreement

The Runciman Royal Commission on Criminal Justice recommended in 1993 that only trained and qualified interpreters be used in court. In response the National Register of Public Service Interpreters (NRPSI) was established in 1994. The NRPSI was administered from 1996 by the Institute of Linguists (IOL), in 2000 it became a limited company limited and a subsidiary of the IOL group, and it finally became independent of the IOL in April 2011. An examination for interpreters, the Diploma in Public Service Interpreting (DPSI), was set up in 1994 with legal, medical and community options administered by the IOL.

In 1998 the Trials Issues Group recommended the exclusive use of National Register Interpreters when selecting interpreters for criminal investigations and court proceedings. This was confirmed by the Auld Report in 2001 which also recommended a review of levels of payment “to encourage more and the best qualified to undertake this work”.

The DPSI in Law has become the gold standard in this country as a qualification for interpreting in legal settings. The Metropolitan Police Test for Interpreters, also administered by the IOL, which focuses mostly on police work and is not an accredited qualification, also allows entry to the National Register.

The European Legal Interpreters and Translators Association (EULITA) has held up the National Register and its use in the Criminal Justice System (CJS) as a beacon of good practice to other countries in the European Union.

Problems with the National Agreement

In Februay 2010 the MOJ established the Interpreting Project to look at reforming the provision of Interpreting and Translation services. Its purpose was to improve the existing system and find ways to reduce costs while maintaining appropriate quality standards. The review identified “three major concerns about existing arrangements “: (Equality Impact Assessment Initial 30.3.11.pdf—Page 3)

(1)The time consuming and costly mechanism for booking interpreters.

(2)The limited availability of registered interpreters in some languages and in some parts of the country.

(3)That complaints made against interpreters are not being investigated … thoroughly....

Some of the factors causing these concerns were:

Access to the NRPSI list of interpreters was limited by the NRPSI board to organisations in the Justice Sector that were prepared to pay a substantial fee for access. Besides the cost this meant some parts of the CJS did not have access to the register.

The search facilities on the NRPSI list were extremely poor so police forces and courts sometimes thought a suitable interpreter was not available when they in fact were.

Police forces and courts sometimes had to make multiple phone calls to find an available interpreter as there was no nationally co-ordinated booking system.

The fees for travel time and parking for each job were calculated individually which was administratively time consuming.

Payments for interpreting jobs were made individually and sometimes still by cheque which was administratively expensive.

NRPSI interpreters were at loggerheads with the NRPSI board which they felt was not running the NRPSI in the best interest of interpreters or of the Justice system. The NRPSI board was responsible for investigating complaints about interpreters. (The structure of the NRPSI has subsequently been reformed and the NRPSI list of interpreters is publically available with improved search facilities).

Besides identifying these concerns the Interpreting Project effectively raised a number of questions:

(1)Was the current pool of NRPSI interpreters sufficient to meet demand or did the pool need to be widened?

(2)Could a wider geographic pool of interpreters decrease costs by reducing travel expenses?

(3)Could people with language skills but without the DPSI, and the training in interpreting skills and legal vocabulary that required, interpret to a sufficient standard in some CJS settings?

(4)Could using such less well trained interpreters working at a reduced rate help to lower costs?

(5)Could a new test sponsored by the chosen agency adequately identify these interpreters?

During its “competitive dialogue” to find a single provider the Interpreting Project was persuaded by the bidding companies or others that the answers to questions 2,3,4,5 above were “yes”. From this was born the concept of Tiers of interpreters where Tier 1 has the DPSI or Met test, Tier 2a has a pass in the legal DPSI apart from the written exam, Tier 2b has an English related degree or equivalent diploma plus some experience of criminal justice work (though not necessarily in this country), and Tier 3 had previous public sector experience. Tiers 1 & 2 also have to have at least 100hrs of public service interpreting experience. All tiers are have to have a pass at the relevant level in the new test.

However no trial of the level of the interpreting skills needed for different CJS jobs was undertaken, nor was a trial undertaken of how well results in the new test, combined with the Tiering requirements, correlated in practice with the interpreting skills needed for courts and police stations. (The new test has subsequently turned out to be mostly a test of an interpreter’s facility with non-legal English with only a few legal terms included and has no in-depth assessment of legal terminology or interpreting skills as the DPSI does)

The Rational for a Framework Agreement for a Single National Contractor for Interpreting and Translation Services

The original aim of the review of interpreting and translation services was to make savings of 10% on an approximate £60 million spend. (e-mail by Richard Mason to Interpreters 30 March 2011). This was a reasonable target but has been side-lined following Applied Language Solutions’ (ALS) promises of savings of £18 million. It is not clear exactly what the much quoted £18 million savings figure applies to, whether the five year contract with a stated total worth of £125 million or to some other figure.

One can see that at the time a Framework Agreement with a single provider seemed like a good solution for the MOJ. Though negotiation with interpreters organisations might have produced the 10% reduction in costs that was being aimed for it would not have dealt with the problems associated with booking interpreters or with complaints. A Framework Agreement delegated the responsibility for the organisation of interpreters, negotiating reduced terms and ensuring adequate standards to a private company and out of the hands of the MOJ. It appeared to also offer a solution to the difficulties with bookings and complaints while adding provision for continuous professional development and appraisal and also promised to save large amounts of money. A single contractor gave the selling point of a single contact phone number, though for a service not used by the public this would seem to have little advantage. Ironically in its tender ALS charges extra for phone bookings to discourage them.

There were however potential problems and downsides to this decision as interpreters pointed out in their responses:

All the large agencies in England had in reality up till then done most of their interpreting work in one or two regions in the country. This meant national roll out for a single contractor was always going to be difficult as local contacts are vital for interpreting agencies.

The MOJ, as a virtual monopoly employer of interpreters used in a legal setting (over 90% at a rough estimate), has a strong motivation to develop a sustainable, high quality service interpreting service. In contrast an agency on a fixed term contract is naturally going to look at maximising its short term profits and to be less concerned with issues of long term standards or provision. This is particularly true for a single provider where there is no competition to drive up standards.

Though a single provider would administratively be much more cost efficient at booking and paying interpreters the MOJ was paying by assignment not in bulk. In order to maintain accounting controls the courts and police would therefore still have considerable administrative costs keeping detailed records of all interpreting assignments and checking the agency’s invoices against their own records.

The Government had recently stopped employing freelance consultants via agencies as it was so much more expensive than employing them direct. There were concerns this considerable extra tier of expense would equally well apply to freelance interpreters, leading to reduced income for interpreters if an agency was involved.

The main evidence that a Framework Agreement might succeed came from the North West Collaborative contract for interpreting which significantly reduced headline interpreter costs for Greater Manchester Police and other North–West forces. This was achieved by ALS paying some NRPSI interpreters £30–35/hour for interpreting time without travel time or travel costs (though its initial rate was lower) and non NRPSI interpreters lower rates. Though the contract had a difficult first two months it succeeded in attracting just enough NRPSI interpreters, supplemented by non-NRPSI interpreters already known to ALS, to fulfil the contract. There were however some particular features which meant that a contract of this sort (ie without or with severely reduced travel time and travel costs) was likely to be difficult to roll out either to courts or nationwide, though at the time some of these reasons were not obvious:

(1)It was on ALS’s “home territory” where ALS already had many non-NRPSI interpreter contacts

(2)The North-West is relatively contained with good motorway connections so travel time was often under two hours.

(3)As an area it contained Manchester, one of the largest cities in the country, so interpreters for most minority languages were near at hand.

(4)It was only for police jobs which, in 90% of cases do not last more than 2hrs, so ALS could sequence jobs relatively easily (courts are much less predictable).

(5)The police are usually willing to wait a few hours for an interpreter while courts often cannot do this. This factor combined with (4) meant that ALS could fulfil the contract with relatively few interpreters

(6)If no travel time is paid the two hour police jobs give a significantly higher hourly rate of pay to interpreters than one hour jobs

(7)Police interviews usually have no witnesses and the North-West police were either not encouraged or actively discouraged from making complaints about any reductions in the standard of interpreting. Therefore though many examples of non-fulfilment and poor interpreting from use of non-NRPSI interpreters were collected by NRPSI interpreters these did not reach the press. The openness of courts to the public and the presence of judges meant courts were unlikely to be so tolerant of reduced standards.

(8)It is not clear if the relatively high rate ALS was paying interpreters, compared to the fee ALS was receiving, was financially sustainable for ALS, as its year end accounts for the year following the contract showed a loss of £331,000.

Difficulties in Drawing up the Contract for the Framework Agreement

The MOJ had minimal information for potential bidders on either the total number of assignments required, or their geographical and language distribution. Companies appear however to have been asked to quote for an hourly rate for each Tier inclusive of travel time and expenses. This method was obviously convenient for the MOJ in calculating fees per case but interpreters had traditionally been paid travel time and travel expenses and these were vital for jobs involving minority languages and geographically isolated sites. As the agencies had no information on these costs their bids were always going to involve a large amount of guesswork. It is reported that ALS, a relative newcomer to the interpreting industry (2003), bid substantially lower than the other two more experienced agencies.

By 30 March 2011 when interpreters were consulted on the new Framework Agreement it was clear that the MOJ was having considerable problems picking out which of the improvements ALS had agreed to were practically achievable and which were not. For instant the key performance indicators in the contract have as their first objective the absolutely impossible target of “availability of all languages within a 25 mile radius—95%” (some languages have less than 10 NRPSI interpreters to cover the country). The MOJ also appears to have had difficulty picking out which of interpreters’ responses to the Agreement contained reasonable points and which did not as this impossible target made its way into the final contract.

The overall impression is that the MOJ placed too much weight on ALS’s sales pitch and either did not seek, or was not able to find, good independent advisors in Public Service Interpreting to guide them in drawing up the Agreement. Knowledgeable and completely independent advisors in this area may in fact not exist but there were knowledgeable people working for the government such as the leads for the Metropolitan Police Interpreting Unit, the Tribunals Interpreting Unit and the Interpreting lead for Cambridgeshire Police. Though none of these three could be considered to be properly independent as they had vested interests in employing interpreters direct rather than through an agency they were still knowledgeable in the field and could have given the MOJ guidance.

Interpreter Pay Under the Framework Agreement and the National Agreement

ALS is offering interpreters £20/hr (assuming most jobs are Tier 2). There is a £5 addition per job which will be more than absorbed by travel expenses for the first 20 miles and so is not included in the calculations below. Additional time is paid per minute. ALS offers travel expenses of 40p/mile after the first 20 miles and travel time after the first 2 hours but no parking fees and these are not included in the calculations of pay below.

Assuming no parking fees, for a one hour job this gives rates of £10/hr, £6.7/hr and £8/hr respectively for 1hr, 2hrs and 4hrs travel time. For a three hour job it is £15/hr, £12/hr and £11.4/hr for 1hr, 2hrs and 4hrs travel time respectively. The average rate of pay working for ALS is therefore around £11/hr.

Under the National Agreement interpreters were paid £30/hr with a minimum interpreting fee of £84 for three hours work. Travel time was paid at £15/hr, mileage at 25 or 40 pence per mile, and parking fees were reimbursed. Assuming no parking fees, and excluding mileage expenses, for a three hour job with one hour travel this gave rates of £25/hr, £23/hr and £20.5/hr for 1hr, 2hrs and 4hrs travel time respectively. This gave an average of around £23/hr.

The Tribunals Service were paying £26/hour and £16/hr travel time with a minimum overall fee of £48 and mileage of 25 pence per mile but not parking. Assuming no parking fees, for a 1hr job this gave rates of £24/hr, £19.3/hr and £18/hr for one hour, two hours and 4 hours travel time respectively. For a three hour job (usually only employment tribunals) this gave £23.5/hr, £22/hr and £20.2/hr for 1 hour, 2 hours and 4 hours travel time respectively. As most jobs are for 1 hour this would give an average of around £20/hr.

It can be seen from the above that the Framework Agreement pays interpreters less than half the hourly rate of the National Agreement and just over half the rate of the Tribunals Service.

It is unclear if the MOJ realised that the rate cut for interpreters would be as dramatic as it is has been as ALS states in its Tender Response (Framework Agreement page 126) that it has always paid market rates and this has been repeated by the MOJ. However careful reading of the Tender Response shows that the way ALS using the term “market rate” is not the usual definition. (see ALS and Probity on page 6).

According to a survey by the Institute of Translation and Interpreting (ITI) in 2011, average annual income for interpreters in this country was £15,000. Interpreters working for the most in demand languages such as Slovak & Romanian can earn around £30,000—£35,000 from the National Agreement and a very small number earn more than this by working very long hours. (The single unnamed interpreter who reported to the press that he earned £100,000 from the National Agreement appears to be mythical). ALS is not likely to require significantly less interpreters than before and an average annual income of £7,500 is simply not sustainable for professionals with postgraduate qualifications.

Problems With Fulfilment of the Framework Agreement

The Framework Agreement has run into significant problems with finding sufficient interpreters to fulfil the jobs required and this has been widely reported in the press. This has been particularly marked in some areas and some languages. http://www.bbc.co.uk/news/uk-17009115, http://www.peterboroughtoday.co.uk/news/local/court_interpreter_service_criticised_1_3577650

http://www.thetelegraphandargus.co.uk/news/local/localbrad/9618688.Bradford_Judge_demands_firm_explain_after_it_failed_to_provide_interpreter_in_case

Both ALS and the MOJ appear to have been taken off guard by the level of resistance to the Framework Agreement by existing NRPSI interpreters. However given that their rate of pay had been halved and that in Dec 2011, once the new rates were known, over 60% of NRPSI interpreters declared that they would not work for ALS these difficulties are not so surprising. Though the trial roll out to the courts in the North-West went smoothly this was in ALS’s “home territory” where they had a large number of existing interpreter contacts. ALS was also able to supplement any shortfall in interpreters by bringing them in from elsewhere in the country.

The level of non-fulfilment since national roll-out has not been made public but from press reports it would appear to be substantial. It is also not clear if the MOJ is collecting its own data on cases where an interpreter fails to show or is relying on data from ALS. Given the questions about ALS’s use of numbers (see page 6) it would seem highly advisable for the MOJ to collect its own data as well if it wishes to get a true picture of the problem.

The only data released so far to give a view of the situation nationwide is a CrimeLine survey of solicitors in the week March 12–16, 10 weeks into the contract, which showed from 403 responses that in 56% of cases the interpreter did not turn up at all, in 18% the interpreter turned up late and in 26% the interpreter turned up on time.

dl.dropbox.com/u/50165963/int…

The cost to the MOJ over the last 2.5 months arising from interpreters not turning up has not been made public. However the MOJ must have estimates of the average cost of a no-show by a defendant in Magistrate’s courts, Crown courts and Tribunals and will therefore have been able to monitor the considerable accumulating costs incurred by the non-fulfilment of the Framework Agreement. There is an automatic nominal penalty fee in the contract for failure to fulfil each job below 98% fulfilment but the cost to the MOJ of a no-show must far outweigh this fee.

The recent shortage of interpreters for court cases has made it clear that though interpreting fees are a minor part of total court costs, in cases where they are needed, interpreters are the oil which enables the machinery of justice to turn and without them it grinds to a halt

The Cost Implications of Using Less Highly Trained Interpreters

There have been numerous reports in the press and on the interpreters’ website www.linguistlounge.org of sub-standard interpreting under the new Framework Agreement. Unless significant numbers of extra NRPSI interpreters sign up in the next few weeks this problem is likely to continue long term even if fulfilment rates gradually edge up.

Assuming most court work has so far been designated as Tier 2 it is likely that already a high percentage of all court and police interpreting jobs are now being done by Tier 2 interpreters ie without the legal DPSI exam. If interpreters have a Tier 2 pass at the new assessment centres, they should have passable skills in interpreting non legal English but from report the test includes minimal legal terminology. They will not therefore have the depth of knowledge of legal terminology or the interpreting skill that candidates for the DPSI have studied hard to achieve. Being able to interpret a series of technical legal terms accurately and at speed is a vital part of being a competent court or police interpreter. Tier 2a interpreters who have failed only the written paper in the DPSI law exam are likely to have a reasonable knowledge of legal terminology but will only make up a small number of total Tier 2 interpreters.

One of the strongest arguments against paying less for lower quality interpreting is that in the end it will lead to much greater expense for the MOJ. The cost-effective running of the justice system is highly dependent on the quality of communication between the parties involved. The Attorney General highlighted last year the huge savings to be made when defendants who are going to plead guilty do so at the earliest opportunity. Doing this depends on high quality and accurate communication between the defendant, the police and the defendant’s solicitor. Sending an untrained interpreter with inadequate knowledge of legal terminology, and without the calm professionalism that comes with higher training, in order to save £11/hr will not prove cost effective given the very high costs of a late guilty plea.

Lower quality interpreting will also increase costs from misinterpretation leading to appeals and miscarriages of justice requiring retrials. As well as being more accurate and providing more effective communication trained interpreters are almost always faster than untrained ones. When conversations are being interpreted consecutively this can make a considerable difference to the time a case takes. Again adding an extra five–10 minutes per hour to each court case by using a lower quality interpreter in order to save £11/hr of interpreter time cannot be cost effective for the MOJ.

Probity and ALS

NRPSI interpreters have for some time been reluctant to work for ALS due to their perceived poor levels of probity. This has been manifest in the past by: their failure on occasions to pay interpreters the agreed rate for jobs; excessively late payments; reneging on orally or previously agreed terms; and misrepresenting probable arrival time with police stations.

The last 2.5 months of the Framework Agreement have shown further examples of dubious probity:

Inflating interpreter numbers by including interpreters for whom they had details but who had not agreed to work for them. In its Tender Response ALS states that “approximately 2,500 of our 4,500 registered freelance interpreters are suitably experienced and qualified for Authority assignments” (page 114). What ALS fails to mention is that this number is almost entirely made up of the database of NRPSI interpreters which it had purchased. ALS would have been quite aware that most of those interpreters neither had any idea they were on their database nor had agreed to work for them. In early March ALS’s number of interpreters dipped suddenly after someone showed that their pet rabbit was registered on ALS’s website suggesting they were including all who had registered at interest, but not necessarily agreed to work for them, in their figures.

CRB checks—on page 149 of the Language Services Framework Contract ALS states: “for all CJS work interpreters are enhanced CRB checked and then vetted…” and that “following recruitment interpreters are subject to the application of the agreed security standards before being made available…for assignments”. However they have relied on new interpreters uploading their own previous CRB check onto their website and from a recent email from Gavin Wheeldon to all ALS’s interpreters encouraging them to do this it appears that quite a number have so far failed to do this.

www.linguistlounge.org search on “CRB”. See reports 4,9,10,11,12 (may shift with more reports)

Tier 2 infringements—Tier 2b interpreters (ie interpreters without the DPSI in law or a fail in only the written exam) are required to have an English related language degree or diploma + proved 100 hours of public sector interpreting experience + previous experience of the justice sector. It is appears from reports of the inexperience of many of the new interpreters that some have been interpreting without one or even any of these requirements and ALS has sometimes relied simply on their Tier test.

www.linguistlounge.org—search on “assessment”. See reports 27 + 5,9,24,25,26,29,30 (may shift)

If this variation on the contract conditions has not been negotiated with the MOJ it would represent a clear breach of the contract conditions. Either way it would go against the MOJ’s assertion that the Framework Agreement is not compromising the quality of interpreting. Since under the Framework Agreement the identity of all interpreters along with their qualifications has to be e-mailed to courts and police stations prior to the interpreter arriving this should be possible to check by checking with courts on the details given for new interpreters who do not have either the DPSI or a fail in the written exam.

Data protection infringements—interpreters who asked for all their data to be removed from the ALS website have recently found this has not been done and even bank details have been kept. A number of complaints with documented evidence of this have been submitted to the Information Commissioner’s Office. http://www.bbc.co.uk/news/uk-england-london-17463036

Market rates—both ALS and the MOJ have repeatedly asserted that under the Framework Agreement they are paying interpreters “market rates”. In its tender response ALS gives as one of its unique selling points that “before and throughout the economic crisis we have continued to pay our linguists above the market rate for their services.” The official definition of “market rate” is “the usual rate in the market”. However when using the term neither ALS or the MOJ have explained that in the Tender Response (Framework Agreement page 126) the phrase does not have its usual meaning but means the rate that ALS’s Interpreter Working Group, a group of interpreters working for them, decide it should be. A request last year, by a private individual for one hour of a Polish NRPSI interpreter’s time, gave quotes from Prestige Network and the Big Word, two of the largest agencies in the North-West, of £150 and £120+mileage respectively, whilst ALS quoted £45 (all excl. VAT).

The Future for the Framework Agreement

From the above it is clear that the average £11/hour paid by ALS to interpreters under the Framework Agreement is an inadequate rate of pay for freelance professionals with postgraduate training. This rate of pay is simply not capable of sustaining the high quality interpreting service that the MOJ requires.

At this stage although a few more NRPSI interpreters may sign up with ALS, the majority of those who have not done so are unlikely to as the pay rates are so poor. ALS’s options for finding more interpreters in this country who meet the tier 2 requirements are by now limited and it is also likely that a proportion of those who have so far signed up will either prove to be inadequately trained for the job or find the work too stressful. High profile cases criticising untrained interprets are also going to cause some Tier 2 interpreters to stop working for ALS. http://www.bbc.co.uk/news/uk-england-london-17709440

Though ALS is now advertising in Eastern European countries this is unlikely to provide the quality or number of interpreters required.

In some languages there is already a severe shortage of interpreters due to reduced numbers of NRPSI interpreters. Given the low pay rates it is very difficult to see how this could be remedied even in the longer term, particularly for less common languages. The MOJ will therefore inevitably end up with a long term supply problem for some languages fairly soon.

Even at this early stage in the Framework Agreement it is likely that a majority of court and police work in this country is now being done by non NRPSI interpreters and it is inevitable over time that standards will fall even further. NRPSI interpreters currently signed up will find in time they can earn better money elsewhere and will gradually leave. As Tier 2 interpreters will be eligible for almost all the work available it will remove any motivation for new interpreters to improve their legal interpreting skills up to DPSI exam level. The exam also costs almost £500 to sit. Though ALS says in its tender response said that it would sponsor linguists to sit the DPSI for rare languages it will not have the money, or the linguists the motivation, to make this happen on a larger scale. Similarly though ALS says it will arrange post-graduate training, with the grade of linguist the pay rates are likely to attract, this is not likely to have much effect. Within a few years it is likely there will be very few NRPSI grade interpreters left working for the CJS.

It is also not clear to what extent the Framework Agreement is still a financially viable proposition for ALS and its owner Capita. It is likely that in order to win the contract and the monopoly this gives that ALS bid on tight margins—reports are that the other more experienced companies bid at a considerably higher hourly rate. ALS has so far officially increased its rates by £5 per job and increased travel costs. However unofficially there are reports that it has been paying many individual interpreters far more than the official rate, either as their language is in short supply, or to encourage interpreters to take jobs with more travel time which are otherwise financially unattractive. http://www.eadt.co.uk/news/ipswich_it_s_a_farce_fed_up_defence_solicitors_blast_court_interpreter_system_after_translator_has_to_travel_from_newcastle_to_ipswich_for_hearing_1_1331803

The automatic penalty for non-fulfilment will also eat into their margins as will solicitors taking out wasted costs orders. These factors combined with the fact ALS as a company made losses of 20,000 in 2010 and 331,000 in 2011 raise questions as to the National Agreement’s financial viability for ALS/Capita in the long term.

Other Models for Saving Money on Interpreter Costs

When the Framework Agreement was negotiated with ALS there was some evidence from the North-West Collaborative contract used by local police forces that by using an agency such contracts could significantly reduce costs while maintaining standards. However information collected at the time by interpreters suggested that standards were variable depending on the training of the interpreter. These problems with standards did not come to light because police interviews have no independent witnesses, and complaints from officers were not encouraged. NHS Manchester which used the same North-West Collaborative Commercial Agency Agreement for non-NRPSI interpreters found that complaints about poor interpreting standards escalated rapidly once it joined the Agreement as the interpreters sent by the agencies (including ALS) were of a lower quality than they had previously been used to. Since then NHS Manchester has started, wherever possible, using its own vetted team of interpreters and has found this gives better standards and is cheaper than using the North-West Collaborative agreement.

Other large employers such as Leeds Hospitals have also found they could reduce costs by not using an agency. http://www.bbc.co.uk/news/uk-england-16868007

Katrina Mayfield the interpreter manager for Cambridgeshire police force reduced costs by 40% last year by actively managing the service. The Metropolitan Police Interpreting Unit and the previous Tribunals Agency Interpreting Unit also run or ran efficient operations at lower cost than other courts or police forces by actively managing their interpreters.

Employing NRPSI Interpreters Direct and Also Achieving Savings

By learning lessons from the Framework Agreement it would be possible for the MOJ to employ NRPSI interpreters direct, at their previous hourly rate, but at a substantially reduced cost to the MOJ. This would give the MOJ savings approaching those of the Framework Agreement but also ensure continued quality.

The three hour minimum fee is appropriate for cases where a court or police station needs to guarantee the presence of an interpreter for three hours or more. However ALS has found that 80% of police jobs take 1.5—2hrs with only 10% taking longer. A routine two hour minimum might therefore be agreed with interpreters for police jobs, with the understanding that interpreters could leave if necessary after two hours, and that at booking the police could request a 3 hour minimum for the occasional case where this seemed necessary. This measure alone would give an approximate 21% reduction in cost for all police work compared to the previous National Agreement. The cost to the MOJ of this Alternative Agreement (AA) would be roughly the same as under the Framework Agreement (excluding admin) on an assumption that the MOJ is paying ALS £45/hr for Tier 2 work.

eg if one assumes the MOJ is paying ALS £45/hr under the Framework Agreement (FWA), a two hour police job would cost £90 (+ the vetting fee). Assuming two hours travel each way the cost under the Alternative Arrangement (AA) given above would be £60 interpreting time + £30 travel time = £90 (+ travel expenses) and under the old National Agreement (NA) £114. Travel time of one hour would make the job cheaper using the Alternative Arrangement than the Framework Agreement—£75(AA) vs £90(FWA) vs £99(NA) and travel time of three hours slightly more expensive £105(AA) vs £90(FWA) vs £129(NA).

Similarly if there are certain court or tribunal cases where the finish time could be guaranteed to be I hour or 2 hours after commencement a system to pay a similarly reduced minimum time might be negotiated. For a court jobs lasting 3 hours the cost to the MOJ with two hours travel time + parking + travel expenses would be roughly the same (£90 + £30 + £15 = £135) under the Alternative Agreement as that paid under the Framework Agreement (£135) and for one hours travelling it would be less (£90 + £15 + £10 = £115). For a one hour job the comparison would be £55(AA), vs £45(FWA) vs £109(NA) for one hour of travel and £60(AA) vs £45(FWA) vs £124(NA) for two hours travel. One hour jobs therefore cost a bit more under the Alternative Agreement than the Framework Agreement but these were particularly badly remunerated using the FWA. Using the AA they still cost half the price of the National Agreement.

Similar payment models to those suggested here already exist in some police forces and in the previous tribunal service booking arrangements. They do reduce the overall hourly rate for interpreters for jobs lasting less than three hours as the balance between travel time and interpreting time is altered. One hour jobs requiring three or more hours travel are worst effected and to compensate for this would need an increase in travel time rate to £16/hr if travel time was three or more hours, and £17/hr if travel time was five or more hours. A similar sort of approach to this used by the previous Tribunals Agency Interpreting Unit with its higher rate than courts for travel to compensate for lower interpreting rates.

The main way these fees reduce costs is by reducing the number of hours interpreters are paid for. Given that these new rates are so much more competitive the MOJ could balance this by instructing that all court and police work should be preferentially booked direct with NRPSI interpreters rather than using agencies. This would compensate NRPSI interpreters with an increased flow of work and save money by raising standards (see above). This combined package might well be acceptable to NRPSI interpreters and would also meet the standard of interpreting recommended by the Runciman and Auld reports. This model for payments would therefore produce considerable savings, not far off those achieved via the Framework Agreement, whilst maintaining the existing skilled workforce.

Making Interpreter Booking and Administration Cheaper and More Efficient

If the MOJ developed a national interpreter booking website on which each court and police station entered details of the cases where an interpreter was required, and each interpreter had a diary page on which they indicated their daily availability, an available interpreter would often be found by a mouse click and a single phone call. Realistic travel time along with parking fees could be calculated automatically, booking emails and invoices generated automatically, and payments automated. A simple grading system for rating the quality of each interpreting job once done could allow courts and police to choose a more highly rated, and usually more experienced, interpreter if wished and also identify any underperforming NRPSI interpreters.

This sort of computerised system would reduce interpreter booking and administration costs to below those that any agency could manage. In particular the MOJ would not have to check and challenge detailed monthly invoices from an agency.

The details of such a system would need to be carefully thought through and set up in negotiation with interpreters and experts in police and court administration. (An addendum to this document gives further detailed suggestions). Initial enquiries to two large and reputable website companies taken out in another context indicate that it should be possible to set up such a website for less than £30,000.

The website would also easily give the MOJ, as well as courts and police, detailed on-going information about their interpreter costs and useful information about the spread of different language groups around the country and how this compared to the spread of interpreters. One of the useful calculations for the website to make would be average travel time for each language and each site along with the number of assignments. If this sort of information were put in the public domain it would allow existing or new interpreters to relocate to under resourced areas if they wished.

Other Suggested Changes to the Organisation of NRPSI Interpreters

Should the above measures be adopted it would still leave some of the organisational matters provided for under the Framework Agreement unaccounted for.

Interpreter management

If the systems of booking and payment above are to work effectively it will require good communication between interpreter bodies and the MOJ and on-going active management by the MOJ. It would therefore be important for the MOJ to appoint an interpreter manager(s) with some knowledge of the field to oversee the system, deal with any on-going issues and liaise with interpreter bodies. Encouraging NRPSI interpreters to form a single representative body would make negotiations between the MOJ and interpreters much simpler. This would also help interpreters as a profession to develop more quickly and consolidate professional standards

Complaints

Less serious problems could be dealt with initially by the Interpreter Manager. More significant complaints should be adequately dealt with by the Disciplinary Committee of the NRPSI now it has been reformed and is independent of other interpreter bodies.

Continuing professional development for interpreters

The professional interpreter bodies such as APCI, ITI, PIA, SOMI etc. could perhaps set up a working group to look at this area and come up with some suggestions of what should be reasonably expected. Starting a system of appraisals for NRPSI interpreters would be best delayed for two–three years until the system is more settled. It is of note that yearly appraisals are not yet mandatory for solicitors though they do have to do CPD.

Other Considerations

Interpreters previously used by the Tribunals Interpreting Unit

There is a cohort of interpreters without the legal DPSI who have for many years given good service to the Tribunals Service. It is suggested a way is found to include them in any new arrangements so they are able to continue doing this sort of work.

Telephone Interpreting

This would need to be put out to tender from an agency to give 24 hour cover. It would be cost-effective to use high quality legally trained telephone interpreters wherever possible and from a reputable agency.

The ALS list of interpreters

It is becoming increasing clear from reports of poor standards of interpreting that the Tiering system in the Framework Agreement does not guarantee sufficient standards and should not be continued. The Met test already gives an opportunity for a legal interpreting qualification for those who feel the DPSI would be too challenging. However the ALS list of Tier 2a interpreters, ie who have passed the DPSI exam except the written component, could be useful to the MOJ, if once the system if up and running there are in practice true shortages of NRPSI interpreters in certain languages.

Addendum A

ADDITIONAL POINTS TO BE CONSIDERED IN CREATING A NATIONAL INTERPRETER BOOKING WEBSITE

Courts and police stations would have a home page with their details of address, contacts, parking rates, entry instructions etc. as well as a bookings spreadsheet.

Interpreters would also each have a secure home page where they could update their own contact details and possibly bank details besides a diary page and a bookings spreadsheet. The diary page would log regular times of unavailability as well as one off occasions.

Court booking pages should allow courts to add any interpreting jobs not booked via the website so that both they and the MOJ have a complete picture of interpreting jobs and costs.

Interpreters’ booking spread sheets should be downloadable to aid completion of tax returns.

The search facility should create a list starting with the nearest interpreter but should also clearly show interpreters ratings so staff could book a more highly rated interpreter if preferred eg for more demanding cases.

The last time and date that an interpreter updated their diary should be clearly shown on the search results list so that booking staff could pick out interpreters who were not actively looking for work.

For jobs likely to last more than three hours courts will need to estimate how long an interpreter will be needed in order for the computerised booking system to know when an interpreter will be available for further work.

All admin such as booking confirmation, agreed terms and costs, and confirmation of completion and payment would all be sent automatically by email. Payments could be aggregated and automated and paid two–four weekly.

Basic travel time could be calculated via a travel website but would need an uplift to fit in with reality and an addition for car park to court time if there was not onsite parking. Travel expenses would have to be mileage costs and assume all journeys were by car. How this level of remuneration worked out for Central London where most interpreters will not go by car would need to be discussed. If courts put the rates for their local car park(s) on their homepage this fee could also be automatically calculated. Finding ways to automate these calculations, even if they are sometimes a little less or more than in reality, is vital for reducing administration costs.

Ratings made by CJS organisations of each interpreting assignment might be on a five point scale such as Very Good, Good, Adequate, Barely Adequate and Inadequate. Allowing the rater to add a comment if wished would be helpful. The rating and comments could automatically appear on each interpreter’s job spread sheet as feedback. A single Inadequate rating or repeated Barely Adequate ratings could create an automatic notification to the Interpreter Manager so he or she could look into the problem. A mechanism for helping those with language deficiencies would need to be devised. More severe or persistent problems could be notified to the NRPSI board.

Annex B

Dear Sir Alan,

I have just read a copy of Peter Handcock’s replies to your questions about the Framework Agreement for Language Services (FWA)sent on 31 May 2012. http://www.publications.parliament.uk/pa/cm201213/cmselect/cmjust/courtinterpreters.pdf

His answers unfortunately illustrate how out of touch those at the highest level in the Civil Service are with the reality of the performance of this contract. I will comment on his individual answers in turn to illustrate this and will send a copy to Peter Handcock as I am sure he would want to be aware of the issues, particularly in light of the Justice Select Committee’s planned investigation.

Introductory Paragraphs

Though the success rate of the FWA is undoubtedly improving and may even in time approach the required 98% the main issue is actually the quality of the service that the MoJ has purchased. When ALS halved the rates of pay of the skilled interpreting workforce, which the MoJ had built up over the past 18 years following the advice of the Runciman Commission and the Auld Report, it was not surprising that the vast majority of those interpreters stopped working for the MoJ.

So far only 301 NRPSI interpreters out of 2,350 previously on the register have signed up with ALS. There are likely to be perhaps another couple of hundred ALS interpreters who have some language training but most linguists will not consider doing demanding part-time freelance interpreting for an average of £11/hr when they can earn more translating or working for other employers.

Fortunately for ALS there are many immigrants in this country who are unemployed or in low paid jobs and speak reasonable conversational English. Though most of these do not comply with the educational and language training standards required for a Tier 2 interpreter, ALS has largely ignored these quality requirements and the MoJ has, so far, not asked the relevant questions. ALS’s “interpreting assessment” has only a handful of legal terms and so provides a convenient and seemingly authorative “Tier 2” label even though it was never intended to be used in this way (see later). There is also considerable doubt, even if all the Tier 2 requirements were fulfilled, whether many Tier 2 interpreters would perform adequately in court as this Tier does not ensure detailed knowledge of legal terminology.

It was inevitable that if allowed to operate in this fashion ALS would eventually find enough “interpreters” , at least in the commoner languages, willing to interpret in courts and tribunals even though in many cases they had no proper understanding of the legal terminology being used.

Britain has a proud history of high standards of justice over many years. Is this the sort of third rate service the MoJ really wishes to provide for those with English as a second language?

Though there is obviously a need to save money on the interpreting budget there are other ways the MoJ can save considerable amounts of money while retaining its trained interpreters. (see the attached Review of the FWA and alternatives, also sent to you previously).

What is the current fulfilment rate? How is this measurement made and verified?

Though the fulfilment rate has increased the complete picture would need to include data on all languages. Even if the contract achieves a success rate of 98% or more in the common languages, if it has a markedly reduced success rate in the less common languages then this will cause a significant long term problem for the MoJ. Though this data is available to the MoJ it was not published in the Statistical Bulletin on 24 May 2012 and parts of the base data, which were published, have been blocked so that no-one else can calculate the figures. This, along with other significant anomalies in the Bulletin, have been discussed in my correspondence with Iain Bell, Chief Statistician at the MoJ, which I have forwarded to you separately. It is not clear whether this data on the “success” figures for all languages has been tabulated and circulated amongst more senior personnel responsible for the contract. This data is important as it is likely to show significant problems with ALS’s success rate with minority languages which it will be difficult to correct even over time.

How many interpreters are now providing services via ALS? How many were providing services under the old arrangements?

The reply says that there are around 1,500 interpreters employed by ALS but that because some speak more than one language this equates to more than 3,000 interpreters by language. ie an increase of 100%. The National Register of Public Service Interpreters (NRPSI) has around 2,350 interpreters but this number only increases by about 20% when extra languages spoken by a single interpreter are included. Relatively few NRPSI interpreters interpret in more than one language because although it is possible to be fluent in ordinary conversational terms in several languages interpreting in court is considerably more demanding than this Far fewer NRPSI interpreters interpret in more than one language because although it is not too difficult to have command of ordinary conversational terms in several languages interpreting in court is considerably more demanding than this. Not only does the interpreter have to learn the equivalent translations for specialised English legal terms, many of which are not in dictionaries, but he or she also needs to be able to simultaneously translate these complex terms at high speed. This means that relatively few people are able to pass the Diploma of Public Service Interpreting (DPSI) in more than one language.

Peter Handcock’s disclosure of the high numbers of ALS interpreters who speak multiple languages is a concrete illustration of the way ALS has focused on finding interpreters for a particular language at the expense of standards. It also illustrates the markedly lower standard of their new assessment test compared to the DPSI, assuming ALS interpreters have taken the assessment in each language.

The reply has also misinterpreted the MoJ’s own Statistical Report as it says that ALS interpreters “cover 142 languages”. The report actually says that “142 languages were requested” but only gives the success rate for the 20 commonest. It is likely that for large number of these 142 languages and probably even the majority, ALS was unable to provide an interpreter. Hopefully the MoJ will now publish data on the success rate for all 142 languages so it is clear to everyone.

ALS should be asked: Of their interpreters who interpret in more than one language what percentage have a separate Tier 2 pass for every extra language marked by Middlesex University?

What qualifications are required of interpreters?

I am interested that Peter Handcock’s reply says that ALS developed the tiered system since I had understood that the tiering system had been developed in conjunction with the MOJ. He also says that each hearing is assigned a tier based on the skills required of the interpreter. In fact despite producing this new system the MoJ & ALS never arranged for an independent assessment to see how Tier 2 interpreters actually performed in different types of court setting so the Tier of interpreter that is required for different types of court work is not known. Furthermore because only 301 of ALS’s 1,500 interpreters have the Diploma of Public Service Interpreting (DPSI) and are therefore Tier 1 it is clear that the default for most court cases has to be for a Tier 2 interpreter even if the court would prefer a Tier 1 grade.

There are numerous anecdotal reports of interpreters with only basic interpreting skills ie Tier 3 on ALS assessment, being offered court work. The following questions need to be asked of ALS to discover if this is correct as Tier 3 interpreters are only intended for community jobs and should not usually be working in courts or tribunals:

1.How many each of Tier 1, 2 and 3 interpreters do you have?

2.How many court or tribunal assignments have been done by Tier 3 interpreters and in what languages?

What checks are made that those qualifications are held?

Peter Handcock quite correctly states that the purpose of the assessment is to check that interpreters have the appropriate skills that match their qualifications. This has been confirmed by Brooke Townsley of Middlesex University who designed the test:

I would emphasise that what was designed at Middlesex University was explicitly NOT intended to mimic or replace the DPSI examination, nor indeed was it designed to serve the same purposes. It was delivered solely as an in-service performance check, providing a diagnostic check (not a pass or fail result) on ability to deliver a simultaneous interpretation of legal discourse delivered in English at 120 wpm into a target language; and the quality of renditions from a the non-English language of the language pair into English, delivered consecutively. It was also designed on the understanding that it would be used to confirm that the skills indicated on paper by existing qualifications were actually evident in performance. http://www.lawgazette.co.uk/blogs/blogs/news-blogs/interpreting-interpreters-strike#comment-14643

This makes it apparent that the assessment is not really testing legal vocabulary in depth and is more to do with a quick assessment of consecutive and simultaneous interpreting ability. It also clear that the qualifications of interpreters, as laid out in the FWA, are vital if the MoJ is to have a chance of providing service which complies with the FWA requirements. Tier 1 interpreters have the DPSI and so can easily be checked on the NRPSI website. Tier 3 interpreters are only intended to be used for community jobs. It is therefore Tier 2 interpreters whose qualifications need the greatest scrutiny. The requirements for these are set out below:

Tier Two

The interpreter must have one or more of the following:

“Partial DPSI” (English Law option) ie the interpreter must have passed all modules with the exception of component 3b (written translation from English);

A degree in linguistics, English philology, Modern Languages or MA in Teaching of English, or other language related diplomas where English figures as part of the course completed.

Together with (in all cases):

Previous or current employment in criminal justice services in their countries of origin, legal training in the UK or abroad, or other exposure to criminal justice work through other channels is also acceptable (volunteer and/or paid work in the community for police services or work for Victim Support, for example).

University level education (any degree).

At least 100 hours public sector interpreting experience;

References; and

A pass at the assessment centre to the tier two standard.

From numerous reports ALS has often relied on the assessment centre test alone to allocate interpreters to Tier 2 and ignored all the other quality requirements. The questions that need to be put to ALS to check on this are:

(1)How many Tier 2 interpreters does ALS have?

(2)How many of its Tier 2 interpreters have a photocopy or uploaded scan (here after called “documentary evidence”) as proof of “Partial DPSI” (English Law option) ie the interpreter must have passed all modules with the exception of component 3b (written translation from English)?

(3)How many of its Tier 2 interpreters without “Partial DPSI” (English Law Option) have documentary evidence of “a degree in linguistics, English philology, Modern Languages or MA in Teaching of English, or other language related diplomas where English figures as part of the course completed.”

(4)For how many of its Tier 2 interpreters does ALS have documentary evidence of a University degree?

(5)For how many of its Tier 2 interpreters does ALS have documentary evidence of at least 100 hours public service experience prior to starting work on the FWA contract?

(6)For how many of their Tier 2 interpreters does ALS have documentary evidence of “previous or current employment in criminal justice services in their countries of origin, legal training in the UK or abroad, or other exposure to criminal justice work through other channels is also acceptable (volunteer and/or paid work in the community for police services or work for Victim Support, for example)”;

(7)For how many of its Tier 2 interpreters does ALS have a documented pass to Tier 2 standard done at the assessment centre and marked by Middlesex University?

The last point may seem obvious but reports of interpreters not receiving their results suggest it is possible that ALS has not paid for some of the interpreting assessments it has had done to be marked.

Once this data has been collected it is highly likely that the MoJ will find that remarkably few of ALS’s tier 2 interpreters fulfil the stated requirements for Tier 2 interpreters and should not therefore be doing this work. If correct this would clearly explain the number of complaints about the standard of interpreting.

CRB Checks

Peter Handcock states that the “ALS Assessment Centres provide the setting to complete all identification and academic document checking as well as initialising the appropriate CRB and vetting processes”. Interpreters going to the centres in the first month or so described taking their documents and no one at the Centres being interested in them. There has also been at least one email from ALS to its interpreters asking them to upload details of their CRB checks if they have not done so. Individual ALS interpreters attending court have also admitted to other interpreters that they have no CRB check. The contract states that all interpreters will have an enhanced CRB check as a minimum requirement. In its Tender Submission (page 150 of the Language Services Framework Contract) ALS states that if an enhanced CRB check is not in place or is more than three years old they will carry out the check. The questions that need to be put to ALS by the MoJ are therefore:

How many interpreters does ALS have available for the MoJ contract?

For how many of these does ALS have a photocopy or uploaded scan of an enhanced CRB check done within the last three years?

How many of these checks were requested after 27 August 2012?

For how many of those with no enhanced CRB check in the last three years do they have a photocopy or uploaded scan of any enhanced CRB check?

In July 2012 a Freedom of Information Act reply from Warwickshire Police Force, who do CRB checks for the ALS contract, showed that since January 2012 ALS made 720 applications for its interpreters for Non Police Personnel Vetting (NPPV 3) status of whom only 80% (574) were accepted. Assuming that only the 300 or so NRPSI interpreters and perhaps an extra 200 extra semi-trained Tier 2 interpreters are likely to have existing CRB checks this could potentially leave a considerable percentage of their interpreters with no enhanced CRB check in the last three years.

Have the qualifications for court interpreters changed since the start of the ALS contract?

Peter Handcock states that “the qualification criteria…….have not been changed in any way as to diminish the importance of the quality and skills required”.

Prior to the ALS contract the accepted qualification for court interpreters was passing the DPSI exam. This consists of an in-depth test of legal terminology using: consecutive and simultaneous interpreting in both language directions; sight translations in both directions and written translations (under exam conditions) in both directions. The ALS Assessment in contrast is booth based and digitally recorded and consists of a single simultaneous interpretation from English into the target language, a single consecutive interpretation from the target language into English, and a written translation which the candidates are allowed to take home. The ALS pieces for interpreting and translation by report contain only a handful of legal terms. The DPSI examination and ALS assessment are therefore obviously not comparable.

Peter Handcock also states that “the requirement has been enhanced …. by the requirement for continuous professional development (CPD).” CPD implies someone already has a professional qualification and is having on-going training and cannot be used to describe initial training workshops for inexperienced new recruits. In the first six months of the contract how many of ALS’s Tier 1 interpreters have attended a training course suitable for CPD, excluding ALS’s initial training workshops?

What monitoring of performance of interpreters is carried out by ALS and HMCTS

Under the details of the contract ALS are meant on a monthly basis to provide the MoJ with the following:

(1)Breakdown of available interpreter numbers by region, language, tier and vetting status.

(2)Details of complaints received, upheld, timescale for resolution and outcome (per collaborative partner).

(3)Number of new interpreters added per tier per region.

(4) Gaps in availability of languages per region

Has this been happening? Does the MoJ know how many Tier 1, 2 and 3 interpreters in each language and in each region ALS has?

How is feedback from sentencers and other court users obtained and acted upon

The fact that logging a complaint against ALS has been restricted to the courts themselves has restricted the ability of other court users to make a complaint. Interpreter organisations have been contacted by solicitors asking how to make a complaint about an ALS interpreter and there is no obvious mechanism for this. The lack of any independent check on complaints has also allowed ALS to manipulate its complaint figures by including complaints about “wrong tier of interpreter” ie poor quality as complaints about “operational issues”. ALS states in its Tender Information (FWA contract page 120) that “no Customer Service Report (re a complaint) will be closed until confirmation is received from the customer that they are happy with the outcome”. Given the very large number of complaints has this been adhered to?

How many interpreters have been dismissed for underperformance since the start of the contract? How does this compare with numbers dismissed under the previous arrangements? What actions are taken in cases of inadequate performance?

Peter Handcock says in his reply that 120 interpreters were contacted to take part in mandatory workshops scheduled during May and that nine interpreters will no longer be contacted by ALS for future work though whether this is over one month or four months is not stated. In relation to comparative figures he says “the total number of interpreters struck off the (NRPSI) register have never been made available to the MOJ”. In fact since the reform of the NRPSI register in April 2011 this information has been available on their website. The reply also forgets that the Tribunals Interpreting Service which has a list of NRPSI and other less trained interpreters will have regularly collected this data. The MoJ need to ask the Tribunals Intepreting Service:

How many interpreters a year, over the last five years, have they had to dismiss due to poor performance?

How many complaints about interpreters have they received per year for the last five years?

What records do you and/or ALS hold on the number of cases which are adjourned or otherwise postponed as a result of inadequate or non-existent interpretation? Do you have an estimate of the costs incurred as a result of such delays? Are comparable figures available for the pre-ALS contract period?

Peter Handcock mentions there is no centrally collected information on adjournments of criminal courts but he fails to mention that the Tribunals Interpreter Service has for many years collected data centrally and, as they have more fixed time slots than courts, there may well also be estimates of the cost of a no-show. They need to be asked:

What percentage of each kind of tribunals were adjourned or postponed each year for the last five years due to inadequate or non-existent interpreting?

What is the estimated cost incurred for each kind of tribunal if it has to be postponed due to a no-show by a party to the tribunal?

From this last question the estimated cost so far of ALS’s interpreters failing to attend tribunals could be easily estimated. The only proviso is that it is possible ALS has manipulated some of the data so that some cases where the interpreter has arrived too late for the case to proceed and which should be labelled “supplier did not attend” have been labelled as “cancelled by customer” (see Reply to Iain Bell sent previously). This can be easily checked by asking the Tribunals Interpreter Service for the percentage of Immigration and Asylum tribunals over the last few years (ie prior to the ALS contract) where the interpreter’s job was cancelled by the Tribunals Service. If this is several percentage points less than 16% it would suggest that cases have been wrongly allocated.

It is not at all clear how comparing the data on cracked trials, as is done in the answer, helps answer the above questions. “Cracked Trial” is defined in the Bulletin on Statistics on the use of Language Services as “cracked trials are usually the result of an acceptable plea being entered by the defendant on the day, or where the prosecution offers no evidence against the defendant”. How this relates to the number of interpreters not turning up is not clear as most trials have been rescheduled if there was no interpreter.

I hope this commentary on Peter Handcock’s answers helps give a fuller picture of how the contract with ALS is actually performing. I also hope that that you will consider requesting the extra information indicated in red from the MoJ as this will give clear evidence as to whether ALS is complying with the quality standards of the contract or not.

Yours sincerely,

Dr Francis Beresford

Annex C

REPLY TO IAIN BELL, CHIEF STATISTICIAN AT THE MINISTRY OF JUSTICE, ABOUT ANOMALIES IN THE PUBLICATION “STATISTICS IN THE USE OF LANGUAGE SERVICES IN COURTS AND TRIBUNALS 30 JANUARY 2012 TO 30 APRIL 2012

18 July 2012

Dear Iain Bell,

Thank you very much for your reply to my queries sent via Farida Uddin. I am particularly pleased that you have answered my email as all the other queries I, and other ordinary members of the public I know, have sent to the MoJ about the Framework Agreement (FWA) in the last eight months have remained unanswered and unacknowledged.

Thank you for the assurance that “Code of practice for Official Statistics” ensures that for official statistics “the public interest should prevail over organisational, political or personal interests” and that “those producing statistical reports are protected from any political pressures that might influence the production or presentation of the statistics”. It may well be that you are not aware of the controversy and difficulties around the FWA and have therefore unwittingly both included highly controversial statements in your report and have also unwittingly not published the data that could be potentially most damaging in assessing the Framework Agreement.

I would like initially therefore to give you some background information that is of relevance to the statistical data and then look in detail at your answers to some of my questions some of which I would like to challenge or otherwise respond to.

You may not be aware but the outsourcing of Language Services has been exceptionally difficult and controversial for the MoJ and has received almost universally negative coverage in the press. Despite widespread criticism from judges, magistrates, barristers and solicitors the official line of the MoJ has been to deny that there are significant problems with the contract or the chosen supplier and that all the difficulties are simply related to the settling in of the new contract. Part of the difficulty for the MoJ has been that it has been very slow to realise that ALS, the relatively small company who won the contract, did not have the levels of probity and corporate governance that the government is used to in larger firms involved in official contracts. In particular their use of words and statistics can be extremely misleading for the unwary.

For example in its Tender Response (page 114 in the Framework Agreement Contract document). ALS states that “approximately 2,500 of our 4,500 registered freelance interpreters are suitably experienced and qualified for Authority assignments”. What ALS fails to mention is that this number is almost entirely made up of the database of NRPSI interpreters which it had purchased and that the vast majority of these 2,500 interpreters had not actually agreed to work for them. This fact became painfully obvious to the MoJ in the first months of the contract. You will note that ALS’s statement is very carefully phrased so as to lead most people into assuming that it has 2,500 interpreters ready to work, while avoiding an outright lie. It may well be there are similar statements in documentation you have received from ALS which you may or may not have picked up on.

For those of your answers that have needed a response I have for simplicity included both my original question (in italics), your answer (non-italics) and my response (italics).

Question 1:

On page 8 of the report you publish details of the Tiers of interpreters in a form which has not previously appeared on official documents. May I ask where the wording for this form of the Tiering requirements came from? Was it provided by the MoJ or ALS? F.B.

Answer

The description of tiers is based upon those which are included within the Framework Agreement. The Framework is available on the Contracts Finder website at:

http://www.contractsfinder.businesslink.gov.uk/Common/View%20Notice.aspx?site=1000&lang=en&noticeid=264052&fs=true I.B.

Reply

For clarification I attach a sheet to this email with the description of tiers of interpreters as it appears in the contract and the description as it appears in your report. You do not actually answer my question as to who wrote the paraphrase you publish. If it comes from your department it would help to remove doubt in the minds of the reader if this were acknowledged.

The paraphrase used in your report completely removes all the precise quality standards required for different levels of interpreters under the contract which are so vital for good interpreting in court. Legal terminology is highly precise, and often carries a wealth of specific detail in a single word so that interpreters with fluent conversational English but no legal training are likely to make major mistakes in interpreting in court. In trying to fulfil the contract ALS has chosen to largely ignore these quality standards as most trained interpreters will not work for them for the pay rates they are offering. Your inclusion therefore of a list of tiers which includes no standards and suspiciously fits with what ALS has been doing in practice, and what they would undoubtedly prefer, is therefore highly contentious! If this is not the official policy of the MoJ it needs to be corrected. F.B.

Question 2:

You say that there have been requests for interpreters in 142 languages but only provide data on the success rate for the twenty most popular. Furthermore many of the entries in the raw data file have the language blocked by “Not disclosed” so it is not possible to produce this additional information from the file. The reason for the “Not disclosed” label is given as “to protect the privacy of individuals” and it has been done for any record where there is a single record for a particular court and particular language in a single month. If this is to protect an interpreter’s identity this makes little sense since knowing that a Latvian interpreter attended Plymouth Magistrates Court in March gives no useful information about the interpreter’s identity. Even if one somehow believed it was true then it is not clear how knowing that a Latvian interpreter attended Plymouth Magistrates Court on two occasions in March discloses any less than knowing that one attended. F.B

Answer

The reason for not disclosing some languages is not to protect the privacy of those who are interpreting. Interpretation is provided for juveniles, vulnerable witnesses and others where we must prevent the inadvertent disclosure of identity. To protect the privacy of interpreters, we have decided not to publish details of individual complaints, which might possibly be linked to an individual. I.B.

Reply

Thank you for letting me know the suppression of language data is to provide inadvertent disclosure of identity for juveniles and vulnerable witnesses. If this is so your method of trying to carry this out unfortunately still makes no sense at all. Your policy means that if Juvenile A appears once with an interpreter in July in a particular court the language will not be disclosed but if he appears twice in July in that court, which could easily happen, then it will be. If you truly believe there is a risk of disclosure by this data, though that seems highly unlikely, the chances of guessing are likely to be the same whether Juvenile A appears once or twice in the court in the month, making your method of protection invalid. F.B.

Question 3

As you may be aware it is for minority languages that ALS is likely to have the lowest success rate, possible 0% for some. This is because for these languages there are usually very few NRPSI interpreters and they would be hard to replace as there is by definition a reduced pool of bilingual speakers. The only reasonable conclusion from the data as presented is that this information is being deliberately withheld as it shows the performance of the contract in a negative light. It is however exactly the sort of information that is needed by the MoJ to judge whether the contract is viable long term long term for all languages. F.B.

Answer

The success rate was published for languages where a reasonable number of requests have been made. This was a decision made purely on statistical grounds, For example, a language with two requests, the success rate can only be 100%, 50% or 0%, none of which is particularly meaningful. It is not good statistical practice to report calculations based on a limited number of observations, which can imply there has been vast improvements or deterioration in performance from one period to the next.

The only reason data has been deliberately withheld has been to protect the identity of individuals. The details of these restrictions have been fully explained in the publications.

The purpose of publishing the statistical bulletin was to inform the public; it does not form part of the ongoing monitoring of ALS’s contract. I.B.

Reply

I am very surprised at your reply suggesting you cannot publish the data on minority languages as there are low numbers of requests, as in your report Table 20 (complaints, page 22) consists mainly of noughts and ones and you presumably considered this to be valid. Furthermore the Northern Ireland Courts and Tribunals service has for some years published data on the number of minority language cases even if there is only one case. http://www.courtsni.gov.uk/en-GB/Services/Interpreteration-Services/Quarterly-Bulletins/Pages/default.aspx . The usual statistical procedure in such situations appears to be to give only absolute numbers and not percentages and to add a caution about interpreting these figures between time periods. This is certainly the approach the EU statistics agency takes in publishing the data for homcides, instead of not including some groups as you have done. http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/Crime_trends_in_detail (see data sources and availability)

Furthermore if the purpose of the statistical bulletin is to inform the public then the information about minority languages is, of all your data, probably of the greatest public importance. For instance, if a Georgian is about to appear in Court he will want to know if a Georgian interpreter will be provided for him. If he sees there have been 3 cases in the last 6 months where a Georgian interpreter was required, none of which have been fulfilled, he may well decide to try and find a knowledgeable friend who can interpret for him rather than endlessly have his case cancelled as has happened to some defendants recently. Having such data in the public domain might also encourage more Georgians to train as interpreters! F.B.

Question 4

Has the success rate for the other 142 languages been calculated and fed back to the MoJ?

Answer

The overall success rate as published in table 1 of the statistical bulletin covers all languages.

Reply

I think you have misunderstood my question. Has the individual success data for each of the 142 languages not mentioned actually been tabulated and if so has this information been passed to the MoJ? If this data has not been calculated by your department it seems unlikely that those in charge of the contract at the MoJ will know either. F.B.

Question 8

Cancelled Customer rates: this is defined on page 25 as “the court or tribunal no longer requires an interpreter and has cancelled the request”. As you may be aware there is no reason in this contract to record prior cancellation as there is no charge for this. There is a charge for cancellation once the interpreter has arrived but this is covered in “DNA customer”. F.B.

Answer

Once a request has been made, it is automatically registered on the management system, and is ready for an interpreter to be assigned to that job. If the court or tribunal cancels the request, the cancellation is also recorded so that an interpreter who may have been assigned to that job knows that it has been cancelled. ALS record information for their internal management purposes which may not be associated with charges made to the department. I.B.

Reply

Thank you. That is helpful in explaining how the system works if the court or tribunal cancels the request and means that at least some of the “Cancelled by customer” entries will be genuine. F.B.

Question 9

In March Cancelled Customer accounted for 18% of all Immigration Tribunals though an experienced interpreter reports that these assignments are almost never cancelled. It has been noticed by observers since Feb 2012 that when court staff phone up ALS because an interpreter has not arrived, ALS give an excuse for the lateness and ask if the court now wishes to cancel (ie as there is no longer sufficient time). It appears that this group represents the cases where court staff say they do wish to cancel. This is partially acknowledged by on page 10 “requests may also fail because …..the interpreter arrives so late the job is cancelled”, though there is no indication where these cases appear. F.B.

Answer

The terms used within the publication are defined in Annex A.’ Annex A makes the meaning of these terms clearer and does not support your interpretation. Any requests that fail because the interpreter fails to attend or is late are counted as “Not fulfilled—Supplier”; any requests that fail because the customer (specifically the person for whom interpretation has been request by the court or tribunal) fails to attend or is late are counted as “Did not attend—Customer”. I.B.

Reply

I entirely agree that one would expect that cases where the interpreter is late would be counted as “Not fulfilled by supplier”. However on careful reading of Annex A, which I assume contains definitions supplied by ALS, I can find no mention of the situation where a request fails as the interpreter is late or of which category it falls into. Under the contract (page 137) cancellation by the customer without charge is generously allowed by ALS up to the point where the interpreter arrives.

From the information I gave above I am still concerned that ALS has inflated the “Cancelled customer” figures by the method above so as to improve its performance and reduce its penalties. The degree to which they would have managed to do this is likely to have varied between courts. As the Tribunals Interpreter Services team previously kept statistics of cancellations, which you will have easy access to, it should be possible to compare the number of cancelled Tribunal cases from 2011 or before with the figures in your report. If the percentage of cancelled cases for Immigration or other Tribunals is significantly more in your report than it has been before, it would, pending any other obvious explanation, support my allegation. This could be checked out further by independent questioning of the clerks at the Immigration Tribunal Courts who were dealing with cases in February and March. F.B.

Question 10

Once these figures are added as “DNA supplier”, where they should be, rather than being excluded, the fulfilment rates over the three months drop to 58%, 73%, 80% (average 72%) ie around 10% worse. This is obviously highly significant because as you may be aware in the contract there are penalties attached to each percentage point less than 98% fulfilment. F.B.

Answer

The bulletin was published to provide the public with information about the use of language services in courts and tribunals, and is independent of the contract monitoring. The success rate as defined in the bulletin is a reasonable measure of the proportion of those requests that could possibly have been fulfilled.

You may also find the following information helpful on processes in place to ensure that the information the MoJ has on language services in court and tribunals is as complete and as accurate as possible:

Court and tribunal staff have been given clear guidance on the circumstances in which cases they should ask for a booking to be cancelled and circumstances in which a booking should be closed unfilled. HMCTS staff and MOJ procurement have full access to the portal to ensure jobs are closed down properly.

The contractor has demonstrated that their staff have a clear set of operating procedures to ensure cases are properly marked as cancelled or unfilled.

Detailed analyses have been conducted by MoJ Contract Managers on cancelled bookings which confirmed that cases are being correctly closed down. I.B.

Reply

Though the bulletin may not officially be part of the contract monitoring exactly the same data will be used for contract monitoring and therefore any alterations to the bulletin data will also alter the contract penalties.

I am pleased that things have now been checked as per your three bullet points. However I am sure that to start with there would have been no procedure for what court staff should do if an interpreter arrived so late that the hearing could not go ahead, as under the old system this was not a significant problem. This, along with the fact that court staff would have been getting used to new procedures and the chaos caused by lack of interpreters, could have easily allowed ALS to misclassify these cases in the first few months of the contract. F.B

Summary of Further Questions Arising From Your Replies

(a)If the controversial and incorrect paraphrasing of the Tiering of interpreters has arisen from your department will you acknowledge this and add a Corrigenda to the report explaining that this is not intended to represent a policy change from the standards for Tiers of interpreters as stated in the original contract?

(b)Given that it is clearly in the public interest, and other reputable official statistics departments have published similar data, will you now publish a table of completed language service requests for all courts and tribunals by outcome for all 142 languages?

(c)Will you check the Cancellation Rate for Immigration and Asylum Tribunals prior to the FWA and compare these with the figures in your report and let the parties below receiving a copy of this letter know the result?

(d)If this Cancellation rate has consistently been more than 2% points lower than the 16% which is the lowest monthly figure so far for Immigration and Asylum Tribunals will you also compare current and past Cancellation rates for other tribunals and request further investigation to check why this has arisen as suggested above? (If the cancellation rate in the past for Immigration and Asylum Tribunals has been 14% or more I will happily accept that there has been no inflation of the figures by ALS).

(e)The other outcome of ALS interpreter assignments that is not covered by your statistics are the not infrequent occasions where the ALS interpreter is unable to stay to complete the case as they have another job to go to and so the case cannot be heard. Can you tell me what protocol has been agreed with court staff and ALS as to how this is classified and entered on the data system and what penalties accrue for such poor performance?

October 2012

Prepared 4th February 2013