Tuesday, March 19, 2019

Horizon trial: day 6 transcript

This is the unperfected transcript of the sixth day of the Horizon trial, held on 19 March 2019 in court 26 of the High Court's Rolls Building. It is the second day of cross-examination of Angela van den Bogerd, a Post Office director. There are a further three Post Office witnesses: Dawn Phillips, Tracy Mather and Paul Smith.

                                         Tuesday, 19 March 2019
   (10.30 am)
                           Housekeeping
   MR DE GARR ROBINSON:  My Lord, redactions.
   MR JUSTICE FRASER:  Yes.
   MR DE GARR ROBINSON:  If I can give your Lordship an update.
       I have been looking at these documents overnight.  So
       far I have found one document where I believe a small
       redaction was mistaken and that's F/1225 and a clean
       copy has been provided to my learned friend, two
       documents where I believe the redactions are correct
       which is F/1549 and F/708.  My Lord, with the others
       I simply need more time.
   MR JUSTICE FRASER:  I didn't intend you to have to do them
       straight away.
   MR DE GARR ROBINSON:  One of them is 109 pages long.
   MR JUSTICE FRASER:  Yes.  So why don't we just leave it that
       there's no need now for you to give me an update until,
       shall we say Thursday morning, and then if you just give
       me a date by which you think you might be able to have
       finished.  I'm not expecting you to have done them
       straight away or on Thursday there will be no criticism
       if you haven't been able to do any more than just these
       three because you are in the middle of calling your
       evidence.
   MR DE GARR ROBINSON:  And it is a slow process.  I have to
       say, on skimming the other documents I have not found
       what appear to be any serious errors that are glaring.
       There are some passages where I would like to speak to
       my instructing solicitors so that I can understand what
       they're about.  They seem to be completely irrelevant
       but I just want to ensure that they are for my own
       satisfaction.
   MR JUSTICE FRASER:  Yes.  I will be taking a slightly
       different approach if I'm told these redactions are made
       on the grounds of relevance rather than made on the
       grounds of legal professional privilege or anything of
       that nature.
   MR DE GARR ROBINSON:  Some of them are on the grounds of
       irrelevance and confidentiality, some of them are on the
       grounds of privilege.
   MR JUSTICE FRASER:  Well, you're going to need to identify
       the ones that are on the grounds of irrelevance because
       I'm likely to be taking a different approach if they are
       said to be irrelevant -- obviously if it's legal
       professional privilege or litigation privilege then
       that's one thing.  If it is said to be on the grounds of
       relevance I will be adopting a slightly different
       approach.
   MR DE GARR ROBINSON:  Well, my Lord, I hear what
       your Lordship says.  It should be understood that the
       Post Office business is massive and extremely diverse,
       so, for example, if there's a review of the car accident
       statistics in a given year, my Lord, that is completely
       irrelevant.  I don't know if your Lordship would expect
       that to be included in any disclosure.  But those are
       the kind of issues we're talking about.
   MR JUSTICE FRASER:  Well, when one begins to dissect
       a document on the basis of parts of that document being
       irrelevant, in the circumstances of this case that's
       something that I can look at and I might choose to look
       at it to decide whether it is in fact irrelevant.  If it
       is legally professionally privileged, I'm not
       obviously -- the assertion of privilege brings that to
       an end.
   MR DE GARR ROBINSON:  Well, my Lord, that would be an
       unusual course for your Lordship to adopt.
   MR JUSTICE FRASER:  It might be, but it's not an improper
       course for me to adopt.  But let's deal with these one
       step at a time, all right?  You have reviewed three, you
       will give me the update on Thursday.  I would like to
       make it clear, though I have already, I don't want to
       impose a burden on you during your evidence of time
       because I have asked you to do it, so if on Thursday all
       you can do is give me an intended timeframe by which the
       exercise might be completed, then that will help.
   MR DE GARR ROBINSON:  Very good, my Lord.
           While we are on the subject of documents there is
       one other point I should mention.  Your Lordship will
       recall the evidence last week, Mr Tank saying that he
       had a box file of documents, and Mr Latif who is still
       a subpostmaster saying that his staff searched records
       which revealed that his transaction correction issue
       occurred in January 2018 not March.  My Lord, my
       understanding is that Mr Tank has now disclosed three
       documents and Mr Latif has disclosed none at all.  Just
       so your Lordship is aware, on Friday my instructing
       solicitors wrote to Freeths making those points and
       seeking disclosure of records and transaction logs
       consulted by Mr Latif for the purpose of preparing his
       witness statement, any notes by Mr Tank, documents in
       the box file and CCTV recordings that were consulted by
       Mr Tank, any other documents that any of the witnesses
       consulted when preparing their witness statements and of
       course any adverse documents.  It sought a response by
       close of business yesterday, we have all been very busy,
       no response was received.  I will keep your Lordship
       informed as to progress on that issue.
   MR JUSTICE FRASER:  Thank you.
   MR GREEN:  I'm grateful, my Lord.  My Lord, I don't know
       whether there's a copy of the unredacted document
       available for your Lordship and the witness.
           (Pause).
   MR JUSTICE FRASER:  Thank you very much.
               MS ANGELA VAN DEN BOGERD (continued)
            Cross-examination by MR GREEN (continued)
   MR GREEN:  Mrs van den Bogerd, we looked briefly at this
       document yesterday.
   MR JUSTICE FRASER:  Is this 1225?
   MR GREEN:  It is 1225, my Lord, yes.
           If we look on page 2 of 4, at the top of it we will
       see there are six bullet points at the top --
   A.  Yes.
   Q.  -- which were previously redacted and now can be seen.
       Do you see those?
   A.  Yes.
   Q.  And those are introduced in fact by the bottom paragraph
       on the previous page, which is page 1 of 4.
   A.  Yes.
   Q.  And that says "Key performance indicators":
           "The programme will track performance against the
       following key performance indicators."
           Yes?
   A.  Yes.
   Q.  And what we see there is that the key performance
       indicators for the programme were reduction of operating
       cost by £3 million a year.
   A.  Yes.
   Q.  Reduction in net agent debt by £1 million.  Reduction in
       subpostmaster suspensions as a result of audit short
       annuals to a level of 60 per year.  Reduction of calls
       into NBSC by 25%.  Reduction of audit losses of 10k or
       over, I think that should be, by 50% and satisfaction
       with online training models of 95%.
           So it's fair so say, isn't it, that one of the key
       thrusts of that KPIs is cost saving?
   A.  Part of it, yes.
   Q.  And indeed that mirrors what we saw yesterday about the
       introduction of Horizon Online?
   A.  Yes partly, yes.
   Q.  And --
   A.  Sorry, can I just say on this particular one the cost
       reduction will come as if -- so the whole thrust of this
       is if we eliminate the errors at the front end then we
       won't need to obviously deal with errors at the back end
       and that's obviously what the cost saving is driven
       from.
   Q.  And so you could see when putting the branch support
       programme together that it was likely to reduce
       subpostmaster suspensions as a result of audit
       shortages, couldn't you?
   A.  Yes.
   Q.  So the measures that you had in mind would be likely to
       lead to reduction in those suspensions?
   A.  Yes.
   Q.  And in a reduction of audit losses of £10,000 or more?
   A.  Yes.
   Q.  And to greatly increased satisfaction in online training
       models?
   A.  Yes.
   Q.  And that reflected a recognition by you of where the
       scope for improvement lay after you had carried out sort
       of horizon scanning exercise, no pun intended, of the
       difficulties that were being encountered by branches?
   A.  Yes.  As I said yesterday it was across a number of
       areas including what was coming through Business as
       Usual, what I saw from the mediation scheme and the
       branch support programme; all of those.
   Q.  Thank you.  Could you now be shown very kindly {F/1330}.
       This is a spreadsheet of call logs that you have
       referred to in your witness statement.  I will just read
       out your witness statement, paragraph 184 {E2/5/42}, to
       you while that's downloading.  Under the "NBSC" heading
       at paragraph 184:
           "In terms of the use of the Helpline and NBSC
       generally, my colleague, Dean Whitehead, WFM & Telephony
       Manager ..."
           What's WFM?
   A.  Workforce management.
   Q.  Workforce management and telephony manager:
           "... has prepared the attached spreadsheet which
       appears ..."
           We can see it is {F/1330}:
           "... which includes the NBSC call data volumes from
       30 March 2015 to 11 November 2018."
           And the aggregated volumes, I understand, were not
       tracked prior to March 2015, is that right?
   A.  I'm not sure it was in this format actually.  We always
       recorded the number of calls but how they have been
       presented has changed.
   Q.  Let's just look then -- in the light of that answer
       let's look at {E2/5/43}, which is what you say at 185.5.
       So you say:
           "In terms of the period of the available call volume
       data, to confirm, while Post Office hold records of the
       calls received that predate 30 March 2015 ..."
           Because we have seen some in this trial, haven't we?
   A.  Yes.
   Q.  "... we do not hold in a readily accessible place any
       aggregated call volume data from before this date."
   A.  Yes.
   Q.  So you would actually have to look and see what the
       numbers of calls were?
   A.  Yes.
   Q.  And then give evidence about that if you wanted to
       explain it?
   A.  Yes.
   Q.  If we can go back please to {F/1330}, let's just look at
       the headings at the top of the columns if we may.  We've
       got A is obviously date, B is "Calls offered" which is
       calls made to the NBSC, C is calls answered, which is
       calls answered by the NBSC, and then D is forecasted
       calls and E is the difference between calls offered and
       called answered.  So that's the number of people not
       getting through, isn't it?
   A.  Hanging up or -- yes, yes, or --
   Q.  Not getting through?
   A.  Yes.
   Q.  They make a call and the call is not actually answered?
   A.  Or they don't progress, yes, through ...
   MR JUSTICE FRASER:  For whatever reason, they don't get
       through to the person.
   MR GREEN:  For whatever reason.
   A.  Yes, because sometimes -- sorry, there's an IVR message
       on the answer, so if you pick up the phone and it will
       say "If you're ringing about this" -- so say we had
       a known problem in the network, we would put a message
       on to the helpline that would say "If you're ringing
       about this particular issue we are aware of it, we're
       looking at it and we will update as soon as we can", so
       at that point someone might drop off.
   Q.  So might these numbers be where there were known
       problems on the network?  Might some of these drop-offs
       be people who had been affected by known problems on the
       network?
   A.  Sometimes, or it could be that they would ring and they
       decide that they don't want to continue to wait, it
       might be that a customer comes in whilst they are on the
       phone and therefore they serve the customer rather than
       waiting on the call.
   Q.  But one way or another they don't actually get through
       to somebody on the helpline?
   A.  Yes, absolutely, that's correct yes.
   MR JUSTICE FRASER:  Mr Green, is this going to be of
       enormous assistance to me in resolving the issues?
   MR GREEN:  My Lord, it probably is of small relevance but
       I have pretty much finished my examination.
           Was there any increase in budget for the helpline in
       the branch support programme?
   A.  In the branch support programme?  I don't ... I don't
       recall an increase in budget.  What we did is we
       restructured so that we had more focus on the tier 2
       support, so that when the calls came in if there were
       calls that couldn't be easily rectified at the T1, that
       get to T2 and we made Horace information available to
       tier 2 as well so they could look into the data in
       branches to be able to assist with problems.
   Q.  So previously, prior to the branch support programme,
       tier 2 couldn't look at the Credence data?
   A.  That's correct.
   Q.  Let's move please to the section in your witness
       statement where you deal with the subpostmaster evidence
       and if we look at paragraph 30 please {E2/5/11}, you
       explain there:
           "To prepare this evidence, I have had to rely on
       a small team of people at Post Office (predominantly
       former trainers and auditors who have complemented their
       hands-on branch knowledge with investigative skills
       honed from several years of investigating claims of
       accounting issues in branches) to help me investigate
       the allegations raised."
           So that sounds like a pretty impressive group of
       people skills wise, is that fair?
   A.  They are good, yes.
   Q.  As you have described them there?
   A.  Yes, yes.
   Q.  And how many people have been in this small team?
   A.  About ten.
   Q.  Ten.  And they are all people at Post Office?
   A.  Yes.
   Q.  And can you remember who they are?
   A.  Yes.
   Q.  Could you tell us who they are?
   A.  So there's Kath Alexander, Shirley Hailstones,
       Reanne Fox, Michelle Cohone(?), Karen Derb(?),
       Peter Todd, Paul Smith, Keith Scott, Matt Waller --
       sorry, how many is that?  I'm trying to think if I have
       forgotten anybody.
   Q.  Well, at least those?
   A.  Yes.
   Q.  If you remember any more you can tell us.
   MR JUSTICE FRASER:  That was nine.
   A.  That was nine.  I'm in trouble now because I have
       forgotten one, haven't I?
   MR GREEN:  That's not something to be in trouble over.
           You say that they have assisted with the preparation
       of your evidence responding to the subpostmaster's
       evidence?
   A.  That's right, yes.
           Sorry, one more was Hugh Williams.  Sorry, it has
       just come back to me.
   Q.  And that's bringing to bear some fairly high-powered
       resources on this point, isn't it?
   A.  Yes.
   Q.  Perhaps more high-powered resources and care and time
       than you might normally be able to devote to someone who
       just raised a query in the normal course of business?
   A.  Well, actually some of them -- that's their role in
       Business as Usual as well.
   Q.  Okay, so they've got experience of actually doing it in
       real cases?
   A.  Yes.
   Q.  And when you looked at this witness statement, this
       section of your witness statement in particular, and
       there were documents that were referred to, did you
       actually look at those documents to see if they did
       support or could conceivably support what you were
       saying in your witness statement about them?
   A.  Yes.
   Q.  You did?
   A.  So they prepared the information and I have reviewed
       everything and looked into that to see whether that
       supports, yes.  I can't recall every single piece of
       information, but that's what I have done.
   Q.  And this has been prepared with the benefit of the ARQ
       data?
   A.  Yes.
   Q.  And the NBSC helpline call logs where available?
   A.  Yes.
   Q.  And the helpline logs would not normally be available to
       an SPM if they called up themselves?
   A.  They are not routinely provided, no.
   Q.  After a call they don't get sent the record --
   A.  No, they don't.
   Q.  -- to see whether it has been correctly recorded or
       fairly reflects what they have rung up about?
   A.  No, that's correct.
   Q.  In fact we see Angela Burke later on has to make a data
       subject access request --
   A.  Yes.
   Q.  -- to get hold of a transcript of her own call.
   A.  Yes.
   Q.  And pay the fee and produce ID in order to get the
       transcript of her own call with the helpline.  Yes?
   A.  Yes.
   Q.  Just parenthetically, do you think that's a satisfactory
       way to respond to a subpostmaster or subpostmistress
       asking for a transcript that you hold of their call?
   A.  No, I think we should be providing it.  If it is -- so
       if we were to routinely provide every call that would be
       quite an onerous task as you would imagine with the
       number of calls.  If it is to support a particular issue
       where there is an issue with, you know, a loss or
       something then I think we should provide.
   Q.  That hasn't been the practice thus far as we have seen?
   A.  No, it hasn't.
   Q.  And let's turn, if we may, to Mr Latif's case which --
       if we look please at paragraph 90 {E2/5/23}.  You say:
           "My strong belief is that Mr Latif has recalled
       these events incorrectly.  The transaction data for the
       branch ..."
           You give three POL numbers:
           "... shows that in the relevant period (June
       to August) every transfer out has a corresponding
       transfer in.  I also note that the ARQ data show that
       there is no stock unit SJ1 as asserted by Mr Latif,
       however there is a stock unit SP1."
           Let's take that in stages.  He fairly accepted there
       was a typo on SJ1 and SP1 so let's leave that aside.
       Now, you understand, do you, the different types of data
       that you can have, transaction data, event data and
       session data?
   A.  Yes.
   Q.  And do you understand the difference between them or
       not?
   A.  From this -- yes, yes, I do, yes.
   Q.  Did you know the difference between them before you did
       this exercise?
   A.  Yes, I have seen the data before.  We have used this
       before.
   Q.  But did you have a really clear understanding of the
       difference between event data and transaction data?
   A.  So the transaction data is everything that -- all the
       transactions.  The event data are things that would
       record transfer in, rem ins, it was the other
       information.
   Q.  What about session data?
   A.  Session data referred to the sessions on there, yes.
   Q.  Do you know really what that means very well, or ..?
   A.  I've not -- in terms of getting into the detail of it
       then not at that level.
   Q.  No.
   A.  What I look at is the information that would be able to
       walk me through a sequence of events and getting under
       the skin of what we can see, put the picture together.
   Q.  Is that level of knowledge why the three POL documents
       you refer to in your witness statement were in fact
       event data not transaction data?
   A.  Sorry, the transaction data should be in there as well.
   Q.  Well, yes, it should, but the three specific document
       references that we had to go and read and look at to try
       and find what you were saying, there's no row or
       anything like that mentioned, those documents are event
       data documents, aren't they?
   A.  Sorry -- well, I understood them to be transactions as
       well.
   Q.  Let's have a look and just trace it through.  Let's look
       at {F/1354} please as an example.  This is the second of
       the POL references in that paragraph.  If we could go to
       sheet 1 rather than the summary please.  So this is the
       event data.
   A.  Yes.
   Q.  It's not the transaction data and it's not data from
       which we can see that every transfer out has
       a corresponding transfer in of the same amount, is it?
       Do you want to go across to the right a bit?
   A.  Not on this you won't see that, no.
   MR JUSTICE FRASER:  Could you keep your voice up please.
   A.  Sorry, yes.
   MR GREEN:  And it shows receipts printed and things like
       that, it shows reversals, variance checks, things like
       that.
   A.  Yes.
   Q.  So you can't actually see that?
   A.  Not from this one, no.
   Q.  No.  And all three of the specific POL numbers,
       documents, at paragraph 90 -- go back to page 23 of your
       witness statement {E2/5/23} -- were actually references
       to events data.  Were you aware that this was taken up
       with your solicitors in February?
   A.  No, I wasn't.  Because to look at this I have looked at
       the event data, as you say, to look at corresponding
       transfers in and out and then the transaction data to
       look at the values of those to see from which stock
       unit, so ... sorry, I didn't put the reference numbers
       on the document, so ...
   Q.  But the POL number documents that were in the statement
       that you had in front of you when you were being shown
       this, I asked you at the beginning did you look at those
       documents to see whether they could support what you
       were saying they showed and you said yes.
   A.  Because I --
   Q.  So did you have other documents in front of you, or ...?
   A.  So I understood that to have included the transaction
       data as well, so that's my mistake in terms of the
       number in there.
   Q.  Very briefly I will just show you this: {H/186/2}.  This
       is a 4 February 2019 letter and it relates to, at the
       bottom of the page, Mr Latif.  Paragraph 8:
           "Mrs van den Bogerd refers to 'the ARQ data' at
       paragraph 90, without exhibit.  Please identify the
       intended document.
           "Also at paragraph 90, Mrs van den Bogerd exhibits
       transactional data for June to August 2015.  We
       anticipate that data outside of this range was
       identified by Post Office and considered by ... her
       small team ... As Post Office must recognise, Mr Latif
       may have misremembered the date of the incident."
           Go over the page please {H/186/3} and then there's
       a reference to data in paragraph 98, checking that, do
       you see that?
   A.  Sorry, where are we?
   Q.  Paragraph 11?
   A.  Yes.
   Q.  The reference to the POL document: is this the correct
       document or not?  And if we look at {H/196/18} there's
       a response on 11 February and the reference to ARQ data
       at the bottom, in paragraph 90 without exhibit says:
           "The reference to 'ARQ data' is to the transaction
       data for the branch which referred to the exhibit
       mentioned in the previous sentence."
           So it specifically asked about is this the
       transaction data in these three event data logs and then
       that comes back.  Were you aware of any of this going on
       in the background in relation to what you had said in
       your witness statement?
   A.  No.  I understood the transaction data was there.
   Q.  Okay.  Because it was actually disclosed on 21 February
       and Mr Latif went to Kashmir on 19 February.
   A.  Right.
   Q.  So if we go back to paragraph 90 {E2/5/23}, you will see
       that the three additional references underneath at
       {F/1353.1}, {F/1365.1} and {F/1371.1}, are additional
       documents which are not those POL documents, but you
       didn't know about any of that?
   A.  No, sorry, I didn't know about that.
   Q.  To find the information about the value of stock
       transfers, which was what was in issue --
   A.  Yes.
   Q.  -- we need to look at the transaction data, don't we?
   A.  Yes, yes.
   Q.  Pausing there, were you aware that there had been two
       quite widespread and persisting bugs which related to
       transfers between stock units going awry when you made
       your statement?  Did anyone tell you that?
   A.  At this time?
   Q.  When you made your statement did you know that within
       the Horizon system there had been two quite important
       bugs which related to problems with stock unit
       transfers, or not?
   A.  Not specifically.
   Q.  No.  Because both the Callendar Square bug and the
       Dalmellington bug related to problems arising on
       transfers between stock units, didn't they?
   A.  I knew about the Dalmellington one, that's an
       outreach -- an outreach issue, so I was aware of the
       Dalmellington.  I hadn't heard about the
       Callendar Square one.
   Q.  But you didn't have the overlap with that in mind when
       you were putting your witness statement together about
       Mr Latif's situation?
   A.  So, no, when I put this together we have looked at the
       data to see whether that -- does the data support what
       Mr Latif has said and what we can see and what we can
       see in there is that we've got the information that
       shows that those transfers have been transferred out and
       transferred in within the branch.
   Q.  Well, let's look at it quickly if we may.
   MR JUSTICE FRASER:  Just before we do I would just like to
       understand your last answer.  Mr Green was asking you if
       you had the overlap between Dalmellington bug and
       Callendar Square bug in your mind when you signed or
       prepared your witness statement dealing with Mr Latif.
   A.  Not specifically.
   MR JUSTICE FRASER:  You didn't, all right.
           Mr Green.
   MR GREEN:  Thank you very much.
           Could we look please now at {F/1365.1}.  And can we
       look at sheet 1 please rather than the summary.  There
       are 28,890 rows.  You didn't refer to any particular row
       number in your witness statement because when you made
       your witness statement you weren't looking at this
       document, or weren't referred to this document, which
       had a different POL number.  Can you remember roughly
       where it was?
   A.  Sorry, where what was?
   Q.  Where you were looking.  Did you do it by date?
   A.  So I would have filtered on it by date.
   Q.  You think you would have done that?
   A.  By value.
   Q.  Did you have it in paper form --
   A.  By stock unit.
   Q.  -- or electronic?
   A.  Electronic.
   Q.  Can we go to row 3478 please.  So there we've got SP1,
       4 July 2015, 13.16.04.  Then 2,000 in column L.  Do you
       want to just quickly see what column L is?
   A.  If you can go to the top that would be the value, yes.
   Q.  So column L is "Sale value", can you see?
   A.  Yes.
   Q.  Column K is quantity, column J is "Product number".  Do
       you see that?
   A.  Yes.
   Q.  If we click back on 3478, so that's on 4 July, £2,000,
       and a minus £2,000 underneath it.  Is that what you were
       looking at, or not?
   A.  Specifically that particular row, is that what you mean?
       Or -- because this is the data, yes.
   Q.  So you filtered by date so we have gone to 4 July,
       beginning of July.  Where did you look?
   A.  So for the whole of July filtered on the value of 2,000
       which would obviously condense the spreadsheet.
   Q.  Okay, do you want to do that?
   A.  Yes.
   Q.  The short point is you can't tell -- let's just pause
       before we do that exercise.  The short point is you
       can't tell from this line what that is, can you?
   A.  Not on this particular line.  So what -- so using the
       event data and the transaction data, looking at the
       event data for the record of the transfers in and
       transfers out, that would tell you all the transfers in
       and the transfers out that had happened in that
       particular -- however you wanted to filter it.  So,
       for instance -- so we looked at June, July and August.
       That would filter it, looking then -- going from the
       event data then to the transaction log for the
       corresponding times and then filter it by stock unit and
       user to narrow it down.
   Q.  Okay.  Do you know what the code 4910 is in that line?
       Do you know what product code that relates to?
   A.  Sorry, I would need that list at the side of me to --
       I'm not that close to the detail.  So there's a list of
       all the product codes and the IDs that we have disclosed
       and I would -- my team would know that off the top of
       their head, I would need to work off the sheet.
   Q.  You would need to look at the ID list.  We can see
       what's been disclosed at {F/1292.2}.  Those are the
       product ID lists, aren't they?
   A.  Yes.
   Q.  And one of the things we notice about the product ID
       lists is they are different lengths, aren't they?
   A.  Yes.
   Q.  So Royal Mail label balance is 32072, that's five
       digits, and, for example, Visa debit payment is also
       five digits, but some other things, stamps and things,
       are only two digits?
   A.  Yes.
   Q.  And they vary, don't they?  We've got four digits for
       Christmas second stamp books.
   A.  Yes, they do.
   Q.  They vary in length.  And some of those product codes
       are provided by third parties, aren't they, so
       Post Office is not in control of them?
   A.  Yes.  I don't know where they come from.
   Q.  Okay.
           Let's go back please -- let's just look, if we may,
       at your statement at page {E2/5/23}, paragraph 91.2.
       That was the analysis that you did in your witness
       statement when you did it in November.
   A.  Yes.
   Q.  And you say there:
           "... there were two transfers of £2,000 from the AA
       stock unit to the SP1 stock unit, both of which were
       successful."
   A.  Yes.
   Q.  And then you say:
           "There were 4 separate transfers of £2,000
       in August 2015 (on the 1st, 3rd, 5th, 26th) all of which
       were successfully transferred into stock unit SP1."
           So that's what you said in --
   A.  November.
   Q.  -- November.  Then you have had to amend your account of
       what transactions in fact took place, haven't you?
   A.  Yes.
   Q.  So 91.2 should now read that in July 2015 there were two
       transfers of £2,000 "between" instead of "from", yes?
   A.  Yes.
   MR JUSTICE FRASER:  We are still in paragraph 90?
   MR GREEN:  We are still in 91.2, my Lord.
   MR JUSTICE FRASER:  91.2, yes.
   MR GREEN:  So it is "between" instead of "from" and "and"
       instead of "to" and then after stock unit effectively
       saying one each way.
   A.  Yes.
   Q.  Both of which were successful, and then 91.3 has also
       been changed to say that instead of four separate
       transfers, in total there were five, adding at the end
       of the sentence "and one on 3 August which was
       successfully transferred into stock unit AA".
   A.  That's right.
   Q.  That correction reflects the fact that the account that
       you had previously given wasn't actually right, was it?
   A.  It wasn't as -- it didn't have as full an explanation as
       I have done on my correction, no.
   Q.  That's slightly finessing it on the first one, isn't it,
       because have a look.  You say there were two transfers
       from AA to SP1.
   A.  Yes.
   Q.  And that wasn't correct, was it?
   A.  No.
   Q.  It's not a question of being full or not full, it's just
       wrong, isn't it?
   A.  Which is why I made the correction, yes.
   Q.  But it is wrong, it's not a question of full or not
       full, is it?
   A.  The statement that said that the transactions -- there
       were corresponding entries for transfers in and out for
       2,000 is right and what I have done in the correction
       is, particularly on 91.3, brought in the other one as
       well and expanded the other one, yes.  So I wanted to
       make it clear on there that there was more than I'd put
       in my statement.
   Q.  Well, I understand adding in the fifth one, but just the
       very short point is 91.2 was wrong, the statement was
       wrong.  They weren't two transfers from AA to SP1 at
       all, were they?  It was one way --
   A.  And one the other.
   Q.  -- and one the other?
   A.  Yes, that's correct.
   Q.  Right.  And you would agree that getting these things
       correct would be very important for fairly resolving any
       issue about discrepancies raised by an SPM?
   A.  Yes, of course.
   Q.  And whether or not the transaction correction process is
       a countermeasure within the Horizon system --
   A.  Yes.
   Q.  -- as Dr Worden has suggested, you are not suggesting
       that this trial is part of the Horizon system and the
       countermeasure, are you?
   A.  Sorry, can you repeat that?
   Q.  Dr Worden regards transaction corrections as one of the
       countermeasures in the robustness of Horizon.
   A.  Yes.
   Q.  It would be are ridiculous to suggest this trial is
       a countermeasure for the robustness of Horizon, so the
       things we find out here are wrong and then can be
       corrected?
   A.  I see, yes, agree.
   Q.  That's an absurd suggestion, isn't it?
   A.  I agree.
   Q.  So let's look, if we may please, at the transaction data
       itself, which is {F/1365.1}.  And let's look at 21 July.
       Parenthetically while that is loading, you didn't
       actually refer to or exhibit the product list, product
       ID list, did you, in your statement?
   A.  No, I didn't.
   Q.  That was necessary to try and decipher these
       spreadsheets?
   A.  Yes.
   Q.  So that had to be found in the disclosure, yes?
   A.  Yes.
   Q.  So having done that let's go please to row 18358 please.
       Now, is this the row that you think is a relevant row
       for this, or is it the next one?
   A.  Sorry, relevant for ..?
   Q.  Relevant for Mr Latif's transfers --
   A.  The transfers in and out from AA to SP1?
   Q.  Which ones should the court look at?
   A.  So it would be both.
   Q.  Okay, do you know what they are both doing?
   A.  It's moving the cash out of --
   Q.  Which one is moving the cash?
   A.  So the first one is 13.54, that's the first one, and the
       corresponding one is the second one below it then.
   Q.  Okay.
   A.  So it happens in two stages on this report.
   Q.  So when Mr Latif was being cross-examined on this -- and
       the transcript reference is Day 2/63 lines 1 to 14
       {Day2/63:1}, we don't need to go there -- it was
       suggested to him that the transactions either side of
       these two entries were back office entries that he
       didn't need to worry about.  Actually, code 1 is cash,
       isn't it?
   A.  That's right.
   Q.  So that's an important part of the transaction?
   A.  Yes.
   Q.  But you would need the product list to know that,
       wouldn't you?
   A.  You would, yes.
   Q.  And 6276 means a transfer in, doesn't it?
   A.  Yes.
   Q.  And 6277 means a transfer out?
   A.  Yes.
   Q.  Are you just being agreeable, or is that something you
       know?
   A.  I would need -- well from this, yes, I know that, but as
       I said earlier, the list is what I would need --
   Q.  If you take it from me at the moment.
   A.  Yes.
   Q.  Us having found the list, that does appear to be what it
       says.
   A.  Yes.
   Q.  And how much of this did you know when you signed your
       witness statement?  Or is this something that you have
       learned more about since you suggested the corrections
       in preparing for the trial?
   A.  So as I said in my witness statement, because of the
       time to put that together I relied on the team to pull
       it all together.  I had reviewed it but not into this
       level of absolute detail.  So what I have looked at is
       the list of the product IDs, that list is something that
       I would need to use as I do use -- because I don't do
       the same frequency as my team, is something I would use
       and that's when I have made the corrections to say
       actually I need to explain that a bit more, which is why
       I made the corrections.
   Q.  Let's compare the transaction data with the event data
       for a second.  Now, if we just look just a little bit
       higher up, can you see 6277 in row 18352?
   A.  Yes.
   Q.  6277, with a 1 above it, that's cash and a transfer out,
       yes?
   A.  Yes.
   Q.  So this sequence appears to begin at 13.52.17, do you
       see that?
   A.  Yes.
   Q.  And Horizon automatically prints out one receipt for
       a transfer in and one receipt for a transfer out,
       doesn't it?
   A.  Yes.
   Q.  I think you mentioned that at paragraph 88 of your
       witness statement.
   A.  Yes.
   Q.  And indeed I think that -- it's probably uncontroversial
       but I can take you to a reference of the original design
       of Horizon, but that's how it was designed to work from
       the outset?
   A.  Yes.  And with a place to sign on the piece of paper as
       well, yes.
   Q.  Can we go please now to the event data at {F/1354},
       which was the second of the original POL reference
       documents you gave in your witness statement, and let's
       look please at row 8528.  Do you see that?  "AP branch
       receipt"?
   A.  Yes.
   Q.  Yes?  And we have then got "Transfer out slip - office
       copy", do you see that?
   A.  Yes.
   Q.  And then at 56134 we've got "AP branch receipt",
       "Transfer in slip - office copy" and "Transfer in slip -
       office copy".
   A.  Sorry, I have lost you.
   MR JUSTICE FRASER:  Use the row number please.
   MR GREEN:  Sorry, 8534.  Do you see 8534?
   A.  The one that is highlighted now, yes.
   Q.  How many transfer in slips do you see printed out there;
       one or two?
   A.  There's one below that as well.
   Q.  Yes.  And they are 4 seconds apart, aren't they?
   A.  Yes.
   Q.  Did you notice that instead of one transfer in slip, two
       had printed out, when you made your witness statement?
   A.  No.  I can't say I did, no.
   Q.  But this was the event data document to which your
       witness statement expressly referred?
   A.  Yes.
   Q.  Two separate receipts being printed out instead of one
       was a feature of the Dalmellington bug, wasn't it?
   A.  I can't recall exactly.
   Q.  Let me show you.  Let's look at {E2/11/26} please.  This
       is in the table annexed to Mr Parker's witness statement
       and do you see the short name is "Unexplained
       discrepancies (duplicate rem in)"; do you see that?
       Under the short name?
   A.  Yes.
   Q.  And there is Mr Coyne's summary and then if we go to
       "Response to Mr Coyne", Fujitsu's comments, if we go
       across to the right, "Financial impact on branch
       accounts":
           "This issue caused a discrepancy in the
       subpostmaster's outreach branch which was easy to
       identify from the transaction logs available through
       Horizon and the fact that separate receipts were printed
       for each transaction."
           Yes?
   A.  Yes, I see that.
   Q.  Did anyone in your small team flag any of this up to you
       to help you give the court a fair picture in your
       witness statement?
   A.  Are you referring to the Dalmellington case or to the
       slips?
   Q.  The fact that there had been two transfer in slips
       printed --
   A.  No.
   Q.  -- and the fact that that was what was being said by
       a witness in this case to be how you could spot the
       Dalmellington bug; did anyone tell you that?
   A.  Sorry, so the Dalmellington bug I knew was specific to
       outreaches.
   Q.  I understand that, but you knew there had been
       problems -- well, you didn't have in the forefront of
       your mind, we have established --
   A.  Yes, yes.
   Q.  -- that these two bugs are about transfers between stock
       units and that was obviously potentially relevant to the
       Mr Latif case.  My question to you is simple: did anyone
       point out to you the two transfer in slips and you have
       said no.
   A.  No.
   Q.  And did anyone point out to you that where there had
       been previous issues with transfers between stock units,
       known problems, one of the indicators was a duplicate
       slip being produced?  No one pointed that out to you,
       did they?
   A.  Not -- no, they didn't, but the Dalmellington is an
       outreach, which is a different scenario to within branch
       transfer anyway, because it's a rem out and rem in in
       the Dalmellington case.
   Q.  Does any of that cause you to wish to think again about
       what you have said about Mr Latif's situation?  Does it
       worry you at all?
   A.  So it is something that I would want to go back and have
       a look at.
   Q.  I understand.
   A.  But if we go back to -- so in terms of the data that
       I have looked at, and I have looked at Mr Latif's data
       in detail, if you go back to and use the data to look at
       the transfers in and out, look at what Mr Latif has
       said, including what he said last week around the fact
       he did that himself, that doesn't show on the data that
       we've got.
   Q.  Okay, well, let's look -- on your understanding of it?
   A.  Mm-hm.
   Q.  Let's look at paragraph 98, the TC/TA issue.  It is at
       {E2/5/24}.  This is still in Mr Latif's case.
   A.  Yes.
   Q.  "The transaction data [and you give a POL reference]
       including data relating to TCs, shows that the branch
       received two TAs on 18 January 2018."
   A.  Yes.
   Q.  Now, the POL reference given to that event data was
       actually to March 2018 data.  Did you spot that when you
       were compiling your witness statement?
   A.  Sorry, no.  I didn't.  The data I looked at was the
       correct data.
   Q.  Yes.  It has now been hyper-linked.  If we follow the
       link that's now on there, it is to the transaction data
       for January 2018 which is the relevant set of data, yes?
   A.  Yes, which is the data I looked at.
   Q.  Okay, did you actually look at the data yourself at the
       time you did this?
   A.  I had a high level, but I have done more work since
       then, yes.  Which is what I set out in my statement as
       I said.
   Q.  I see.  Let's look please at {F/1761.1} and again
       there's no reference in your witness statement to where
       the data that you are referring to shows what you are
       saying.
   A.  Yes.
   Q.  And you can see why that would be helpful for other
       people to understand what your theory is?
   A.  Yes, these aren't the easiest things to ...
   Q.  Especially difficult if no one tells you which specific
       entries they are talking about?
   A.  Yes, agreed.
   Q.  If we look at row 12983 please.  Can you see there are
       four entries there?  Are those the entries you had
       a look at?
   A.  Sorry, on 12983?
   Q.  Yes, 12983 is the first of four entries.
   A.  Right, yes.
   Q.  Are those the ones you had in mind, or can you not
       really remember?
   A.  It was the 25.
   Q.  Yes.
   A.  Two 25s booked in.
   Q.  There are two 25s there, aren't there?
   A.  Yes.
   Q.  And they are both positive 25s?
   A.  Yes.
   Q.  And do you know what the product codes were?  Did you
       look that up?
   A.  From this I can't recall, but what came in was the
       scratchcards.
   Q.  Okay, well, just to help you, 35341 appears to be £10
       scratchcards?
   A.  Yes.
   Q.  33327 is £3 scratchcards and 5473 is £5 scratchcards.
   A.  So the issue was with the £10 ones.
   Q.  Okay, let's look at that and follow it through.  You say
       in your witness statement that two transaction
       acknowledgements decreased stock when they should have
       increased stock.
   A.  That's correct.
   Q.  Which ones are they?
   A.  25.
   Q.  And if we could look please at line 35341.
   MR JUSTICE FRASER:  Can you use the row not the product
       code.
   MR GREEN:  I'm so sorry.
           Can we just go please to 4 January I think at
       row 2237.  So we've got 2237, 4 January.  I just want to
       show you how this appears to work.  So at 6.55, this is
       the same code, yes, 35341?
   A.  Yes.
   Q.  At 6.55 in the morning you've got minus 20, do you see
       that?
   A.  Yes.
   Q.  If we look at row 2309, at 8.54 in the morning we've got
       plus 20, do you see that?  35341.
   A.  Yes.
   Q.  So you've got minus 20 followed by plus 20.  Let's go to
       row 6046.  Do you see that's 9 January at 6.11 in the
       morning?  You've got again 35341, you've got minus 25.
   A.  Yes.
   Q.  And then if we go to row 6213, at 9.15, do you see there
       you've got plus 25?  And the way we found this was by
       searching on the code -- it is obviously quicker in
       court to go to the specific row numbers, but this is by
       searching the code 35341 and tracing through the
       entries.
           So you get a negative entry and then a corresponding
       positive one?
   A.  Mm-hm.
   Q.  Yes?  And we see it again at row 10202 and there's the
       minus 25 and 10331, plus 25.  Had anyone pointed out
       that pattern to you before you did your witness
       statement?
   A.  Not before I did my witness statement.
   Q.  No.  Has anyone pointed it out since?
   A.  So what I asked and I have seen is that the two TAs that
       came in actually came in with the wrong signage on it,
       so it actually came in and it decreased the stock rather
       than increasing the stock and that's why the TC was sent
       to correct that.
   Q.  Yes, but what I'm asking about first is looking at the
       underlying transactions to see what actually happens
       first.
   A.  Yes.
   Q.  And trace the transaction through from the beginning,
       not just look at the TA.
   A.  Yes.
   Q.  Now, it may not be that this is something you know about
       or understand and that's a perfectly reasonable answer
       to give to the court.  You're not required to answer
       questions you don't know.
   A.  Yes.
   Q.  So is this something that you yourself actually know
       about, or has it come from other people who aren't here?
   A.  So my team are closer to this --
   Q.  I know.
   A.  -- but the working -- so the team do the working of it.
       I have interrogated the information to see if I can see
       exactly how that flows and on this what I can see is
       that two lots of scratchcards, £10, came in and they
       came in the wrong way and it has actually decreased
       instead of increased, therefore they weren't -- they
       weren't effectively remmed in, which is what the TA does
       now, and when you look at the data you can see that
       flowing through and what's happened in branch.
   Q.  Well, let's have a look at it, what does actually happen
       on 18 January and I will suggest something to you and
       you tell his Lordship whether you can agree or disagree
       with it.
           Let's look at row 12983.  We've got two positive 25s
       there, do you see that?
   A.  Yes.
   Q.  And if one of the 25s should have been minus 25 he is
       going to be actually 50 short, isn't he?  Because
       instead of recording the amount of stock he actually
       should have had and cancelling out, he is going to end
       up with a record which is 50 rather than zero and
       therefore he is going to be 50 short, do you agree?
   A.  Right.  What he had was minus though.
   Q.  Well, he's got two positive 25s there, hasn't he?
   A.  Yes.
   Q.  He has not got what we have seen elsewhere, minus
       25/plus 25.  So all I'm pointing out to you is on this
       premise he is going to be 50 short, the different
       between the 25s cancelling each other out and them both
       being the same sign, the difference is not 25, it's 50,
       isn't it?
   A.  Sorry, when you say 50 short, what do you mean, because
       he has already got the scratchcards in the branch?
   MR JUSTICE FRASER:  Okay, let me just try and speed this up.
       I know what your evidence is in totality.  What Mr Green
       is doing is he is taking you through some steps.
   A.  Yes.
   MR JUSTICE FRASER:  Now, you might not agree with the steps,
       which is why he is putting it as a premise, but his most
       recent question was the other examples he had shown you
       had a positive 25 and a negative 25 that had an overall
       effect of zero.
   A.  Yes.
   MR JUSTICE FRASER:  Yes?  And what he is putting to you
       here, which is a hypothesis on his case, is that if one
       of these should have been a minus 25, instead of the two
       25s having a net effect of zero they would have an
       effect of 50.
   A.  Okay.
   MR JUSTICE FRASER:  He is going to reput the question
       because I'm not going to put questions for him, but
       that's what he is trying to do.
   A.  All right.
   MR GREEN:  So if that hypothesis is right, the difference is
       that it will show an additional 50 of stock compared to
       what it would have showed had it netted to zero?
   A.  I think so, yes.
   Q.  And he would then not have the artificially inflated
       stock on this premise?
   A.  Yes.
   Q.  And so he would be 50 short.  Real life --
   A.  Yes.
   Q.  -- would fall short by 50?
   A.  Yes.
   Q.  And he would then regard himself as due a transaction
       correction for 50 cards rather than for any other
       number, yes?
   A.  For 50 cards?
   Q.  Yes.
   A.  Yes.
   Q.  On that premise?
   A.  Yes.
   Q.  And in the event that he reverses a transaction for 50
       cards, his cards should go down by 50?
   A.  Yes.
   Q.  But his cash should go up by 50 correspondingly?
   A.  Yes.
   Q.  And so he would want, in fairness, a transaction
       correction to increase his stock by 50 cards?
   A.  Yes.  And reduce his cash.
   Q.  Yes.  In fact the TC that he was sent increased his
       stock by 100 cards, which caused a 50 card overshoot,
       a 50 £10 card overshoot.  He is therefore left £500
       short on that premise, isn't he?  If that's correct?
   A.  Yes, but that's not -- so if I can just -- can I just
       explain?  So what he should have had was -- so it is two
       25s, £500 worth.  Effectively he has activated the
       scratchcards and therefore what he should have had --
       sorry.  So if you just think about it before we did the
       PING.  So it is basically you activate your scratchcards
       and you get the rem in so it balances.  What came in was
       actually the wrong way round so it didn't take the 500
       to zero, it took it to a minus so it actually is 1,000
       not 500 is the effect.
   Q.  But the difference he is complaining about is 500.
   A.  Yes, why -- okay, the difference is then -- so
       Mr Latif's practice was to -- as soon as he activated
       the scratchcards in his post office he would then buy
       them -- sell them on to his retail.  So what he would do
       straight away is he would process a sale for £500.
       That's what he would do and that was his practice and
       that's quite -- a lot of people do that because it is
       easier to keep the cash separate.
           That would show as a sale on the information.  Now,
       what he has done in this case is when -- because he's
       got negative stock figures, he couldn't balance with the
       negative stock figures and then when they have looked
       they can see that they've got negative stock figure on
       the scratchcard, it's showing a minus.  So they then
       reversed the sale of the 500 but actually they had never
       sold them in the first place, so actually he had reduced
       it.  So what that means is actually the scratchcards
       that were in the retail hadn't been paid for, so there's
       actually £1,000 short is what the TC comes in should do
       and that was correct.
   Q.  Well, let's look on the call logs, {F/1834.1}.  If we
       look at 25 January and we look at row 48.  So row 48 is
       10 August 2015 and you see there:
           "We have received the wrong TAs for £5 cards."
           That's what he is calling the helpline about
       in August 2015?
   A.  Okay.
   Q.  Yes?  And then if we go to row 56, 13 August 2015:
           "Can PM has received TAs this morning and the office
       has been charged for the free games."
           You see that?
   A.  Yes.
   Q.  And then if we look at row 57, 14 August 2015, rollover
       with discrepancy, yes?  Caused by lottery TAs being
       wrong.  This happened quite a lot to him.  Yes?
   A.  Yes.
   Q.  And then row 72, on 25 January 2016, rather than 2018,
       he's got a duplicate lottery TC there as well.
   A.  Yes.
   Q.  Can you see that?
   A.  Mm-hm.
   Q.  So there's a history there of him having these duplicate
       lottery problems and raising them and if we go to
       row 171, 25 January 2018, do you see:
           "TC received for ... £10 x 100 scratchcard
       games ..."
           And if we look at the top, he says:
           "... only received 50 ... if we accept the TC, will
       cause a discrepancy?"
   A.  It is right he had 50 but because of the way the TA was
       sent, it was sent -- it is because it had the complete
       opposite effect it actually went from being 500 to
       1,000.  So it didn't -- so whereas he should have booked
       in 500, he booked in a minus 500 which is the effect of
       £1,000.  That is the difference.
   Q.  He is specifically complaining about the fact that the
       TC that he has received is not to correct the 50, it's
       to correct 100 and if he accepts it it will increase his
       stock by 100 lots of £10 scratchcards.  That's precisely
       the discrepancy that he has complained about in his
       witness statement, isn't it, £500?
   A.  But the value of the TC was 1,000 --
   Q.  Precisely --
   A.  -- which is what it should have been.
   Q.  -- so it overshot by 500 on his case for the reasons we
       have just outlined and put to you?
   A.  No, it didn't though.  Because -- because of the way it
       was -- sorry, I'm not explaining this obviously.
       Because of the way it was booked in, instead of taking
       it to a zero it took it to a minus, so it had the double
       effect of that coming in, so he -- actually he only had
       50 scratchcards, two packs of 25 and £10, only 50, but
       because of the way it came in, instead of increasing it
       decreased so basically instead of going from 500 to zero
       it took it down again, so it's £1,000.  And from what
       you -- you know, there you say he's got a couple -- and
       I have said it clearly is confusing for him in branch
       because of the way that's happened and because of the
       way he actually sells his stock to the retail and
       because of the way they reversed that transaction, it
       wouldn't be particularly easy to follow in branch, but
       that TC was correct.
   MR JUSTICE FRASER:  Mr Green, I am entirely comfortable that
       I can see the degree to which the call log either does
       or doesn't match either Mr Latif's evidence or
       Mrs van den Bogerd's understanding of that.
   MR GREEN:  I'm most grateful, my Lord.  I was going to move
       on anyway.
   MR JUSTICE FRASER:  There is nothing to be gained by
       continuing.
   MR GREEN:  I'm grateful.
           Would that be a convenient moment for a break,
       my Lord?
   MR JUSTICE FRASER:  Yes.  I do have to explore with you your
       intended timings.
   MR GREEN:  My Lord, yes.
   MR JUSTICE FRASER:  You have the whole of this week but
       you're not going to get any more time just because we
       are going slower with one witness than you expected.
   MR GREEN:  No, we're not seeking in any and we're planning
       on that basis.
   MR JUSTICE FRASER:  All right.  Mrs van den Bogerd, we're
       going to have a short break now.  We will come back in
       at 11.53.  Thank you very much.
   (11.45 am)
                          (Short Break)
   (11.54 am)
   MR GREEN:  Mrs van den Bogerd, can we now look please at
       paragraph 77 of your witness statement at {E2/5/20}.
           Now, this was in fact a shortfall raised by you
       rather than raised by Mr Tank, wasn't it?
   A.  Yes.
   Q.  So in a sense it's a slightly random one, it's not one
       that he was specifically complaining about but it's one
       that you identified as having occurred in December 2011
       and you say that he contacted NBSC regarding the
       shortfall, the call was made on the 13th:
           "I understand from Post Office's solicitors and
       Fujitsu that this particular issue was resolved
       following an investigation as PEAK PC0214226.  It is
       possible that Mr Tank contacted the banking team direct
       (if he had their contact details), but such calls are
       not logged."
           And the result of this is that he is repaid.
   A.  Yes.
   Q.  And you say at paragraph 87 you have described above at
       paragraph 53:
           "... the process which subpostmasters need to follow
       when there is a system outage or power failure and,
       providing this process is followed, Horizon will either
       recover or cancel the transaction.  The online banking
       transaction which Mr Tank has described is a recoverable
       transaction and therefore ..."
           This is the important bit:
           "... if the correct recovery process had been
       properly followed in accordance with the Horizon Online
       quick reference guide, a copy of which appears at [POL
       number which you have now amended to the correct one],
       the branch would not have sustained any shortfall."
           So the point there is user error by Mr Tank; that's
       the point you're making?
   A.  So -- no, the point here is that the -- well, there was
       a user error from Mr Tank in that he gave all the
       receipts to the customer.
   Q.  Sorry, say again?
   A.  In that he gave all the receipts to the customer, that
       was his error, so he didn't have that information.  But
       in this -- so there's two things.  In the case of
       Mr Tank then what's happened in his situation is that
       other than give the receipts, didn't do anything -- it
       wasn't user error because the recovery failed and
       actually --
   Q.  Yes.
   A.  Sorry.  And what happened here was that the failed
       recovery was then picked up by Fujitsu in the back end
       that comes in to Post Office and then we generate the TC
       to the branch, which is what actually did happen here.
   Q.  Well, let's see what actually did happen, I mean let's
       just be very precise, because I'm going to suggest to
       you that when anything happens because of Horizon, if
       there is any user error by an SPM that happens at the
       same time but is not causative, Post Office mentions
       user error in a way that suggests it is causative.
       Do you recognise that?
   A.  In this situation, yes.
   Q.  Yes.  And the fact is that any user error by Mr Tank was
       not causative of the error that was generated?
   A.  No, it wasn't.
   Q.  And so when we re-read that part which I said was the
       important bit at the top of page 21, you say:
           "... therefore if the correct recovery process had
       been properly followed in accordance with the
       Horizon Online quick reference guide ... the branch
       would not have sustained any shortfall."
           That's not correct, is it?
   A.  Not in this situation, no.
   Q.  No.
   A.  It was -- so in this situation it was the receipts,
       because on the receipts it actually does tell Mr Tank
       what -- you know, what the situation -- what's happened.
   Q.  Well, let's take it in stages.  Do you know what UEB is
       as an acronym?
   A.  Sorry, U?
   Q.  UEB, have you ever heard of that?
   A.  No.
   Q.  User error bias?
   A.  No, I can't say --
   Q.  It is where people in IT constantly blame the user when
       actually it is not their fault?
   A.  Okay, I haven't heard that before.
   Q.  You haven't heard that?
   A.  No.
   Q.  Well, let's look at this because I'm suggesting to you
       this is a common theme that runs through Post Office's
       approach when these issues are raised.  Is that a fair
       suggestion?
   A.  I think what I have tried to explain is what the
       approach is, in 99 out of 100 that would be the case.
       In this situation it is slightly different.
   Q.  Well, it's slightly different in the sense that it is
       not his fault and you have suggested that it is?
   A.  Right, well, I didn't intend to suggest it was
       completely his fault.  I mean him giving the receipts
       back to the customer was user error and had he held onto
       those he would have seen exactly ... but other than that
       I mean he couldn't have done anything because it was
       a failed recovery anyway and he did do what the receipts
       will have told him which is to pay out, which is what he
       did.
   Q.  He did what the receipts told him to do, didn't he?
   A.  To pay out.  It was an authorised payment, yes.
   Q.  So when we follow it through -- I'm going to have to
       take it reasonably speedily, stop me if there's anything
       that you need more time on.
   A.  Okay.
   Q.  But if we look at {F/870} we can see this is the PEAK
       and if we go to page 2 please {F/870/2} and look at the
       summary, do you see the second box down:
           "The banking transaction had completed (A3 received
       and authorised ...), including the receipt print ... and
       money should have changed hands."
           Yes?
   A.  Yes.
   Q.  So that's what Horizon would be directing him to do on
       the receipt, wouldn't it?
   A.  That's correct.
   Q.  And that is what he did?
   A.  Yes.
   Q.  We then see:
           "The disconnected session receipts show 'cash to
       customer 195.04', so the customer's account should be
       correct but the branch will have a shortage (for
       a withdrawal) because the session hasn't been recorded."
   A.  That's correct.
   Q.  Can we pause there.  Is this Horizon working as it
       should?
   A.  In that it catches it in the back-end, then yes.
   Q.  And there's a reference to a further PEAK I need not
       take you to it, but it's at {F/871.2} which shows us
       that there are in fact 492 failed recovery transactions
       captured in that PEAK.  Were you aware of that?
   A.  I wasn't aware of that number, no.
   Q.  Did anyone show you the KEL at {F/1700}?  Can we look at
       that please.  And you can see the date of this KEL has
       been -- it has been raised by Anne Chambers on
       28 February 2010.
   A.  Yes.
   Q.  Did you know that there was a KEL relating to all this
       that had been going on since 2010?
   A.  I wasn't aware.
   Q.  You weren't aware of that.
           If we just go down please under "problem", just have
       a look there, there's a paragraph underneath the
       subparagraphs (a) to (e):
           "If a T1 recovery request times out at the
       counter ..."
           Because timing out is a problem, isn't it?
   A.  It is if it is mid-transaction, yes.
   Q.  "... recovery is abandoned and no second attempt is made
       to get the recovery information.  This is as designed;
       it was decided to keep recovery simple and not have too
       many error paths.  The priority is to get the user
       working again, so in this sort of error path we just
       mark the recovery as failed and leave it for SSC to sort
       out."
           So it is designed into the system that in that
       situation the SPM will have a discrepancy that they
       can't figure out and has to be corrected by SSC; that's
       the effect of it, isn't it?
   A.  When you say not figure out, the receipts which is -- as
       Mrs Burke's evidence was, she had the receipts and she
       could figure it out, but it would have a discrepancy,
       absolutely.
   Q.  Well, they can't figure out why, when they have done
       what the actual receipt, if they have kept it, tells
       them to do, they have still got a discrepancy?
   A.  No, what I'm saying is when you say "figure out", it
       would be -- as Mrs Burke's evidence was it was obvious
       to her what that problem was because it didn't show on
       the transaction log, so ...
   Q.  That's as designed anyway?
   A.  Yes, as designed, yes.
   Q.  Let's look please now if we may at the forum post which
       relates to this.  It's at {F/1257.1}.  You see:
           "Hi all,
           "Some advice/help required.
           "Yesterday during HOL failure [Horizon Online
       failure] was in process of POCA card withdrawal.
       Transaction seemed to go through okay apart from Horizon
       printing 3 identical receipts.
           "Receipts showed a disconnected session with
       recovery code.
           "Receipts also showed CAWD limit [minus 195.04]
       total due to the customer."
           Et cetera, we have confirmed he is right about that?
   A.  Yes.
   Q.  We have seen it in the PEAK?
   A.  Yes.
   Q.  "Because receipts showed cash due to customer, we paid
       out.
           "Come evening balancing till showed approximately
       £200 loss.  Thought at time must be miscount and will
       try to sort in morning.
           "This morning produced transaction log for the
       period of HOL failure.  No record of 195.04 transaction
       at all!!!!"
           Yes?
   A.  Yes.
   Q.  That's because it is not recorded internally as
       a transaction?
   A.  That's right.
   Q.  And it doesn't show up on the report that the
       subpostmaster has access to?
   A.  No, it doesn't.
   Q.  "Phoned helpline and was told by very irate member of
       staff that loss is mine unless I can sort out with
       customer directly ..."
           In your experience is that the sort of typical
       response in that situation?
   A.  No, and that's the wrong response clearly.
   Q.  It's wrong, but is it typical?
   A.  No, I don't recognise that.  I mean if he has rung up
       and said "I've got a disconnected session receipt" then
       that should trigger a -- clearly not an irate member
       anyway but it would trigger a different response.
   Q.  On a lot of these helpline records we see them say that
       they read out the KBA in response.
   A.  Yes.
   Q.  That's the knowledge based article.
   A.  Yes.
   Q.  Another word for that is a script?
   A.  It would be where the information is, yes.
   Q.  Yes.  So often the helpline are responding by reference
       to something from the knowledge based article?
   A.  Yes.  So rather than put the actual thing in they
       actually refer to knowledge base, yes.
   Q.  Okay.  Then after the brackets where he says "HAS ANYONE
       SEEN THIS" in capitals, he says:
           "Asked irate staff to pass call up as I was not
       a happy bunny, was told she was not going to do this,
       only after I asked to speak to contracts manager or
       somebody from POL press office with regards to speaking
       to press about my loss was I given a number for
       Chesterfield."
           Is that acceptable?
   A.  No.
   Q.  Is it typical?
   A.  No.  I mean you would get the odd, you know, bad
       experience, but that is not typical in my experience of
       helpline staff.
   Q.  "So spoke to POCA lady at Chesterfield who after
       pressing a few buttons was able to find transaction ..."
           So it is clear that Post Office had access to
       information that the SPM couldn't access; that's
       correct, isn't it?
   A.  Would be having, yes.  Yes.
   Q.  "She couldn't promise anything but will see if it she
       can get a credit TC after she has spoken to Fujitsu?????
       She took my [number] and promised to call back after
       speaking to Fujitsu, being very non-committal about the
       possible loss."
           If you go over the page {F/1257.1/2} he is asking
       about what course of action and so forth.
           Now, pausing there, it's clear that he faced
       a pretty serious headwind, didn't he, in trying to sort
       that out?
   A.  From the response from the helpline, yes.
   Q.  Can we just look at page 6 please {F/1257.1/6}.  The
       second entry down:
           "When I balanced on 17th ..."
           This is not him, this is a different user on the
       same forum:
           "When I balanced on 17th September I had an
       unexplained loss of £176.74.  I paid it in rather than
       go through to the stress of the hell line.  It has not
       come to light."
           Had you referred to the helpline as being called the
       hell line?  Had you heard of that?
   A.  I have heard of it, yes, I have.
   Q.  You have heard of it?
   A.  Yes.
   Q.  And you can understand why people faced with that sort
       of response might take that view?
   A.  Well, given that I said it wasn't typical, I think for
       £176 I would always ring, personally.
   Q.  So that tends to suggest that the person there had had
       some pretty bad experiences if they weren't prepared to
       follow that up, because it's quite a lot of money?
   A.  Well, yes, if they were happy to put £176 when they
       could have sought help with it ...
   Q.  Okay.  Let's have a look very quickly please at the
       helpline log at {F/1286.1}.  Because as well as the
       forum post we just looked at and his description of what
       he did the very next day, we have what the helpline
       recorded from his call at row 120.  I'm sorry, on the
       remedy tab, Ms Mackenzie has very helpfully reminded me.
       Row 120.
           Can we just go across please to the detailed
       description:
           "Called this morning about a Horizon failure
       yesterday, branch completed a withdrawal and the 195.04
       failed recovery receipt time ... recovery code ... card
       account withdrawal limit."
           Just go across to the right please.  You see
       "Priority: low" in the incident log, do you see that?
   A.  Yes.
   Q.  Do you think that should be a low priority?
   A.  That's the classification on the remedy system, it
       doesn't actually apply to that particular call, it's
       just a hangover from the old system.
   Q.  What is that recording if not the priority to be
       afforded to that incident?
   A.  So that would -- so my understanding of that is if it
       was a robbery incident it would come as a high.  It
       doesn't mean they're not taking that call seriously at
       all.
   Q.  At the very lowest, we can see the detailed description
       doesn't exactly reflect the conversation that took
       place, does it, putting it mildly?
   A.  No, it's quite specific to the actual issue itself.
   Q.  It doesn't capture what helpline told him, what he had
       to do to insist on trying to get a number for
       Chesterfield, it doesn't capture any of that?
   A.  No, it has the detail of the failed recovery.
   MR JUSTICE FRASER:  Have you finished with this document?
   MR GREEN:  My Lord, yes.
   MR JUSTICE FRASER:  Can we just scroll to the left so I can
       see the first row again please.  Thank you.
   MR GREEN:  Can we now look please at the Horizon Online
       quick reference guide that you have referred to at
       paragraph 78 of your witness statement.  It is at
       {F/1365}.  This is actually the 30 July 2015 one, but
       this is the guide you would expect an SPM to look at,
       yes.
   A.  Yes.
   Q.  And can you see the red box at the bottom?
   A.  Yes.
   Q.  "Total due to customer", yes?
   A.  Yes.
   Q.  And it actually positively says:
           "You must take care to only settle with the customer
       for the amount specified on the receipt which is clearly
       stated as the 'total due to/from customer'."
   A.  Yes.
   Q.  Where do we find there or at all a statement that "We
       know there's a design fault recognised within Horizon
       which may leave you puzzled when you give a customer
       cash as you have been told to do in this paragraph but
       actually the transaction has disappeared, please call
       the helpline"?  Where do we find something or anything
       remotely about that?
   A.  It doesn't say here.
   Q.  It is not in there anywhere, is it?
   A.  No.
   MR JUSTICE FRASER:  I dare say that wasn't the most
       difficult question that Mrs van den Bogerd had to
       answer.  That's really straying into ...
   MR GREEN:  My Lord, I'm sorry.
           It had been a known issue since February 2010 at the
       latest.
   A.  Okay.
   Q.  Let's move now if we may please, just going at some
       speed, just in relation to the label transaction issues
       that Mr Tank experienced.  His evidence was that he
       couldn't use a spoiled label process because the guide
       told him that he couldn't do so unless the label was on
       hand.
   A.  Okay.
   Q.  And if it has not been printed out it's not going to be
       on hand, is it?
   A.  No, but you can print a receipt, that's part of the
       process, and claim on the receipt.
   Q.  Well, he had suffered repeated losses, become very
       frustrated and if we look at {F/1399.1}, this is the
       transcript of a meeting which he has with Mr Bridges and
       when we go please to page 14 of that {F/1399.1/14}, he
       explained openly about his use of official postage, he
       said it was just a form of protest to get this meeting
       and he is asked "So how many forms of protest do you
       need to get a meeting?", "Well apparently it took all of
       this, and however many years ..."
           You were familiar, weren't you, with SPMs who got
       very frustrated over issues they felt were not being
       faced up to, getting frustrated they couldn't get
       face-to-face meetings with people?  He was not alone in
       that, was he?
   A.  Well, I don't have great knowledge of it, but there
       would be some instances where face-to-face would have,
       in my view, resolved the situation much quicker.  As in
       here he said it only took one call and they had the
       meeting.
   Q.  Well, I haven't got time to deal with the long run-up to
       that.
           Let's look at page 16 {F/1399.1/16}.  If you look
       halfway down he says:
           "I need some acknowledgement of the fact that there
       is an issue.  I received a letter back from KB saying
       that there is, he has investigated it and there is
       nothing to report.  Could I see the results of your
       investigation.  I would like to see the screenshots of
       the conversation that I had with various postmasters and
       the business analyst from Post Office Limited on
       Subspace."
           Subspace was an area people could share their
       concerns?
   A.  Yes.
   Q.  And Post Office withdrew it?
   A.  Yes.
   Q.  "So this is a known fault and when I see
       Paula Reynolds ..."
           It must be Vennells, because I think in the other
       transcript we have it is Vennells.
   A.  Yes.
   Q.  "... when I see Paula [Vennells] standing in front of
       politicians saying there is no known fault with Horizon,
       something gets my goat."
           This is typical of the frustration that you have
       encountered from SPMs, isn't it?
   A.  In some instances, yes.
   Q.  Can we move on now please to the Patnys and I'm going to
       have to take this slightly more speedily.
           In relation to paragraph 62 of your witness
       statement {E2/5/18}, the POL reference there, the
       document that actually was referenced, was not the
       helpline logs, it was the helpline service desk logs,
       yes, which is a different document?
   A.  Yes.
   Q.  That's now been corrected in your corrections to the
       correct POL reference 05114278.  Indeed the document at
       paragraph 63 has been corrected on Thursday of last week
       to the correct number so that they match the documents
       to which hyperlinks had been added on Opus.
           Do you know how it was -- we have covered this
       already at the last time of asking, but do you know how
       it was that all these document references were to the
       wrong things?
   A.  Sorry, I don't.  I mean I had the information I used and
       I didn't do the referencing myself, so I just apologise
       for that.
   Q.  If we look at -- there was a request for cash
       declarations made on 4 February 2019 which is at
       {H/186}.  I won't take you to it.  And your solicitors
       wrote back and said, on 11 February at {H/196}, that
       they would be provided.
           What we get at {F/1514.1}, if we can look at that,
       wasn't disclosed until 7 March.  Did you know about any
       of those requests for documents that related to your
       evidence in the background or not?
   A.  No, I haven't been involved in that.
   Q.  Could we just go into the properties of that document
       please, "File" at the top.  We see this document was
       created on 1 March 2019 by somebody called
       Andrew Keighley.
   MR JUSTICE FRASER:  "Properties" I think.  You are being
       asked to look at "Related dates" under "Properties" on
       the right-hand side.  Do you see?  Have you got the
       common screen?
   A.  Oh, sorry.  Right, yes, sorry.
   MR JUSTICE FRASER:  It is an information block that is
       attached to the file and it says "Properties" in the
       right-hand corner.  Do you see that?
   A.  Sorry, I was expecting a spreadsheet, sorry.  Yes.
   MR GREEN:  Do you see it is created on 1 March 2019?
   A.  Yes.
   Q.  And the author is somebody called Andrew Keighley.  Who
       is he?
   A.  He works in the cash management part of the supply
       chain.
   Q.  And Shirley Hailstones is someone who helped you with
       your witness statement in November?
   A.  She is, yes, in my team, yes.
   Q.  So did you have this document, if we go back, or
       a different document when you were looking at the
       Patnys' situation?
   A.  I don't recall which date it was but certainly when
       I looked at it I looked at the transaction log and the
       event logs for this, but not all of this.
   Q.  I'm going to take -- Mrs van den Bogerd, you will
       understand I'm just going to take this a bit more
       speedily.
   A.  Yes.
   Q.  When you were dealing with the analysis of the Patnys'
       situation in paragraph 66 {E2/5/18} and so forth --
   A.  Yes.
   Q.  -- you didn't mention stamps anywhere?
   A.  No.
   Q.  And you didn't mention calls to the NBSC?
   A.  No.
   Q.  Was there in he reason for that?  They are pretty
       prominent features of the case?
   A.  No, other than I said this was at a time where I put
       this together, then I didn't have a huge amount of time
       to put this together, but ...
   Q.  Okay.  What we see in the call logs -- and I'm talking
       about the logs we have I think already seen but they are
       at {F/1509.1} if we could just look at those.  You were
       in court I think when you heard a positive allegation of
       dishonesty being put to Mr Aakash Patny, weren't you?
   A.  Yes, I was.
   Q.  That he was inflating the stamps to conceal cash
       deficit?
   A.  Yes, I was here, yes.
   Q.  But that's not an allegation that you suggested in your
       witness statement?
   A.  No, I haven't.
   Q.  And the reason for that is you don't believe that's
       necessarily what he was doing?
   A.  I didn't offer an opinion at all.  What I was in this
       looking to do was to try and look at what he said had
       happened and look at what we could see from the data.
       Now, what we can see is that the stamps are overdeclared
       and that has a corresponding cash entry and there was
       a loss beforehand, but I haven't gone further with any
       other suggestion on this.
   Q.  Just pausing there, you will remember from his evidence
       that you heard that he explained that when he called the
       helpline they took him through some complicated steps to
       try and sort it out.  Do you remember that?  Do you
       remember that evidence?
   A.  Sorry, I don't actually.  I thought it was when he spoke
       to somebody in Chesterfield rather than the helpline.
   Q.  It may have been in Chesterfield.  In any event he was
       taken -- I think there are a number of occasions, but
       the point is this, that the Credence report would show
       the specific keystrokes by the operator in branch,
       wouldn't they?
   A.  Yes.
   Q.  And so if one had wanted to get to the truth about what
       keystrokes Mr Patny had pressed, one could have obtained
       a Credence report to identify that and that would
       have --
   A.  Have used a Credence report, yes.
   Q.  And that would have shown all the keystrokes pressed --
   A.  Yes.
   Q.  -- so you could follow exactly what he had done in
       sequence?
   A.  Yes.
   Q.  And we can see at row 140 of this spreadsheet, if we go
       across, you will see the £17,000 shortage.  If you just
       look at the top you can see that -- if we go to the
       "Resolution" column which is to the right, and you look
       at line 140, if you go to "Emailed ESG", just click on
       that.  Above can you see:
           "Emailed ESG Thursday 26/05/2016 ... requesting
       Credence report.  Please see attached documents.
       Advised office of Credence info office adamant stamps
       are declaring themselves and overriding [his] figure.
       Referred to IT if believes system issue."
           So at that date it looks as if a Credence report
       should have been requested, doesn't it?
   A.  Yes.
   Q.  And the Credence report would have shown the keystrokes?
   A.  Yes.
   Q.  And if Mr Patny had been dishonest, it would provide
       powerful evidence of his dishonesty, wouldn't it, the
       precise keystrokes entered?
   A.  Potentially, yes.
   Q.  And it is Mr Patny we see chasing up and repeatedly
       requesting the Credence report in these logs?
   A.  Yes.
   Q.  It doesn't make any sense if he thinks he has been
       dishonest, does it?
   A.  Well, he wants to see the information.  I mean that's
       what he is after, isn't he?
   Q.  Do you know why a Credence report would not be produced
       in a situation like this?
   A.  So the Credence -- I mean it's actually not Credence in
       here that's -- so there is some -- so, sorry, the NBSC
       don't use Credence, they use Horace, which gives a --
       Credence only goes back three months, Horace goes back
       six months, which is better --
   Q.  But this is within --
   A.  So we don't normally produce either Credence or Horace
       information for branches themselves because it would be
       difficult for them to interpret the data, as you found
       out looking at the different lines.  But in terms of if
       he has requested and we said we would produce then we
       should have produced.
   Q.  If we look at row -- well, I need not take you to more
       of it.
           My Lord, I think I can deal with some of the
       transaction issues in the spreadsheets in closings
       without putting them to this witness.
           Let's turn now to the MoneyGram issue, if we may,
       paragraph 71 of your witness statement {E2/5/19}.  Again
       the POL reference was to event data not to transaction
       data.  If we go to the correct document which is
       {F/1436.1} please, just trace this through if we may,
       can we look at the sheet 1 and go please to row 1066.
       Do you see that?
   A.  Yes.
   Q.  Do you know what those codes are?
   A.  Not off the top of my head, no.  Other than "1" being
       cash.
   Q.  So it appears when we compare this with the session data
       we can see what they are, but can I just invite you to
       look at those five lines from 1066.
   A.  So it's the MoneyGram for 3,100, yes, I see that.
   Q.  Yes, 3,100, 3,100.  And if we look now -- you see how
       many lines there are there that relate to -- appear to
       relate to the MoneyGram transaction?
   A.  Yes.
   Q.  So we've got five transactions there in that run.  40170
       is MoneyGram repeat send, just to get your eye in.
   MR JUSTICE FRASER:  Code 40170 is MoneyGram ..?
   MR GREEN:  Repeat send.
   MR JUSTICE FRASER:  Yes.
   MR GREEN:  And if we look down you can see 40173 is
       MoneyGram cancel.  So you've got five lines there from
       40170 down to 40173.
           Can we look at the session data now please to match
       that up, it's {F/1437.1}.  If we go to row 4449 please.
       23 February, if we just go across to the right very
       slightly you will see here that we can now see from the
       session data the item alongside the code: 40170.  If we
       click on the MoneyGram I think we can see above
       "MoneyGram repeat send", and we come down one, we've got
       "MoneyGram send classic" and then we've got underneath
       it "Visa debit payment", underneath that we've got
       "Cash" and then underneath that we've got "MoneyGram
       cancel".
           So we've got two transactions at 12.37 recorded, one
       at 12.42 and two at 12.49.  Do you see those?
   A.  Yes.
   Q.  How many session receipts would you expect to find in
       the event data and for which entries?
   A.  I'm not sure.
   Q.  Let's look at the event data, {F/1435}.  It is on the
       sheet 1, sorry.  And if we look at row 429, we have only
       got four session receipts.
   A.  Okay.
   Q.  And if we just look at the time there, the first two are
       the two transactions at 12.49, but there doesn't seem to
       be a session receipt between 12.13, which is up above at
       line 422, and 12.49.  Do you know why there don't seem
       to be any session receipts for the two lines we saw at
       12.37 or the one we saw at 1.42?
   A.  Sorry, I don't know the answer.
   Q.  So paragraph 72 of your witness statement, if we go back
       to that please, it is {E2/5/19}, you have now amended
       that paragraph, essentially deleting all the words
       after -- well, from and including "Mr Patny says" in the
       third line.
   A.  Yes.
   Q.  And the reason for that is because that's precisely what
       we see when we look at the helpline advice, that he was
       told to settle to cash and reverse and what he did was
       settle to cash and not reverse at the time, but he says
       he did so later at just after 7 o'clock in the evening,
       yes?
   A.  He was told to cancel and reverse, yes.
   Q.  Has any of the evidence that I have taken you through
       yesterday about the fact that this is in the week with
       the PEAK latency --
   A.  Right.
   Q.  -- in relation to MoneyGram -- that massive spike in
       latency, do you remember?
   A.  Mm-hm.
   Q.  And the fact that your account of what people were told
       to do in terms of calling NBSC didn't actually apply to
       Mr Patny because it came afterwards -- do you remember
       yesterday?
   A.  The instruction you mean?  Yes.
   Q.  Yes.
   A.  Yes.
   Q.  The instruction was not to call NBSC at the time he did
       it, was it?
   A.  No -- well, yes, just after it came out, yes.
   Q.  So it took place in the week with PEAK latency problems.
       Duplication of transactions was a known issue in
       relation to MoneyGram in that specific week.  He did
       precisely what the helpline told him in terms of
       settling to cash and that's why he did it and Mr Patny
       appeared to be worried about the polling time being
       after 7 o'clock, in circumstances when we saw that
       Post Office's internal documents didn't pick up the
       disconnect and correct it between polling time for
       MoneyGram and Post Office until more than a year after
       this.
           Now, does any of that cause you to reconsider your
       view of what may have happened in Mr Patny's case with
       his difficulties with an apparent duplication of £3,100?
   A.  Sorry, looking at the information that I have seen on
       this, he did ring the helpline, so he did ring anyway.
       Whether he had had the instruction or not, but he knew
       he had a problem so he rang.  I can see no evidence of
       reversal there, which is why we issued the TC.
   Q.  And are you aware that reversals don't sometimes show up
       in the data that's recorded by Post Office?
   A.  So on -- well, we have issued a TC because he hadn't
       reversed, which is why we have issued the TC.  Had he
       reversed and we had issued a TC, he would have had
       a positive variance from that TC.
   Q.  Short point to you: I'm just asking you, he made the
       point that he was actually doing his reversal that he
       had forgotten to do at the time, he accepted that was
       not right.
   A.  Yes.
   Q.  He makes the point that he attempted to do a reversal at
       7.04 in the evening but that was after polling time.
   A.  Okay.
   Q.  You have seen that at the very time he encountered what
       he believed was a doubling problem, internal documents
       show that that was a recognised feature of the latency
       issues which were at their peak in that very week.
   A.  Okay.
   Q.  Does that cause you to reappraise the likelihood of him
       being at fault in the way you have suggested?
   A.  Well, I think at the time if we were aware of that we
       should have looked into that for him when he has rung
       into the helpline, if we were aware of that at the time.
   Q.  No one told him, "Mr Patny, in fairness to you this is
       a system-wide problem which is peaking this week", did
       they?
   A.  No.
   Q.  It's fair to say that Post Office is not very
       forthcoming about the problems that are being
       experienced more widely when SPMs ring up the helpline?
   A.  Had -- so when there is a problem the helpline clearly
       are aware there's a problem because they get inundated
       with calls, so they are aware when there's a problem and
       usually what happens is that if there's a problem they
       will say "We are aware of it and we will come back to
       you" which is some of the evidence we have heard around
       some of the -- so the issue we had on 9 May for
       instance, we heard about that.  So they were aware of it
       and they will say.  So at the time if there was an
       issue, it should have been -- that should have been
       taken into account, yes.
   Q.  Do you accept that Post Office suffers from user error
       bias in the way that it meets concerns raised by SPMs?
   A.  In some instances, yes, I do.
   Q.  Can we look please at an example at {F/1687.1}.  This is
       Mrs Burke, who we now accept was not at fault for
       causing the problem that she experienced, yes?
   A.  Yes.
   Q.  Can you look please at row 5.  You will see there's a TC
       issued.  Can we click in the text for that please.  What
       she is told is:
           "To correct communications failure on 9.5.16 for in
       pounds 150 ... so credit to office.  Financial services
       enquiry team ..."
           See what's said at the end there:
           "Please be careful when entering transactions."
   A.  Yes.
   Q.  Does that reflect a fair assessment of the cause of that
       problem that she encountered?
   A.  No, that's referring to the TC I think there.  It was
       when you put the TC through.
   Q.  She receives that as a description of the TC correcting
       what's gone wrong, which is nothing to do with her.
   A.  Okay, right.
   Q.  And it says at the end of it "Please be careful when
       entering transactions".
   A.  Which is just a standard line on TCs I think.  It's not
       referring -- what I'm reading from that it's not
       referring to Mrs Burke that she needs to be careful.
   Q.  I'm grateful.
           My Lord, unless your Lordship has any questions,
       I have no further --
   MR JUSTICE FRASER:  Mr De Garr Robinson might have some
       re-examination.  I've got a couple of questions but
       I will leave --
   MR DE GARR ROBINSON:  My Lord, I have a number of questions.
   MR JUSTICE FRASER:  I thought you might.
              Re-examination by MR DE GARR ROBINSON
   MR DE GARR ROBINSON:  Mrs van den Bogerd, I would like to
       talk about MoneyGram first because my learned friend has
       just put to you that the very week that this MoneyGram
       problem arose there was a known issue about duplicated
       MoneyGram transactions and it was put to you on the
       basis that what was being experienced in branch
       reflected the known problem and I would like to
       investigate that with you if I may.
           If we could go please to {F/1555}.  You were shown
       this by Mr Green yesterday.  Do you recall seeing it?
   A.  Yes.
   Q.  And could you just describe what this document is?
   A.  So it's a change request -- well, a pricing document for
       work to make a change request.
   Q.  So some work is being done and it is to sort out
       a problem, isn't it?
   A.  Yes.
   Q.  And if I could ask you to go just about halfway down the
       page underneath "Background" -- I will read it out to
       you:
           "MoneyGram summary.
           "For the last several months Post Office has
       experienced a live operational issue with MoneyGram
       transactions across the branch network.  In the event of
       a transaction timing out at the counter, a system error
       message is displayed to the user (error code 84) and the
       transaction is aborted.  This leaves no record of the
       transaction at the counter and the transaction and funds
       may or may not have been committed to the MoneyGram
       domain."
           If this problem arose in a particular branch, what
       impact would the MoneyGram transaction have had in the
       branch's accounts?
   A.  It depends what stage of the transaction it would be.
       So if it had committed and it left no transaction record
       then it would be having a discrepancy.
   Q.  So -- let me just see if it ... well, let me just --
   A.  It depends.  If the transaction is aborted before it is
       actually registered then it would have no impact at all.
   Q.  Thank you.
   A.  It depends on, as I say, what stage of the transaction
       it would be at.
   Q.  But the description -- I'm just asking you about the
       effect of the description that's here, I'm not asking
       you about anything else.
   A.  Okay.
   Q.  And it says that the problem involves the transaction
       being aborted.
   A.  Yes.
   Q.  And there being no record of the transaction at the
       counter.
   A.  Yes.
   Q.  And my question to you is what impact does that have on
       the branch's accounts?
   A.  In this instance it should have none at all.
   Q.  No impact, okay.
   MR JUSTICE FRASER:  When you mean "this instance" do you
       mean this specific one --
   A.  Because this is a particular -- sorry.
   MR JUSTICE FRASER:  Well, it goes on to say "the
       transactional funds may or may not have been committed
       in the domain".
   A.  Yes, and the point is -- reading the first error code my
       understanding of this is -- and that transaction is
       aborted, it would have no impact at all because it would
       have been cancelled mid-air almost.
   MR JUSTICE FRASER:  Understood.
   A.  If a transaction had already been committed then unless
       it was cancelled and reversed, as we have been
       discussing, then it would almost have been as if it had
       been paid out, so that would have been an issue.
   MR JUSTICE FRASER:  That's what I understood the case to be.
   MR DE GARR ROBINSON:  Well, what I'm seeking to explore with
       you, Mrs van den Bogerd, is first of all what happens
       with the branch accounts and secondly what happens with
       MoneyGram in accordance with the description here.
   A.  Yes.
   Q.  It says:
           "... funds may or may not have been committed in the
       MoneyGram domain."
           So what would happen to the MoneyGram accounts if
       funds had been committed to the MoneyGram domain?
   A.  Sorry, can you repeat that?
   Q.  We have a transaction done at branch which is aborted.
   A.  Yes.
   Q.  And we have no record of the transaction at the counter.
   A.  Yes.
   Q.  But we have funds being committed in the MoneyGram
       domain.
   A.  Right.
   Q.  So we have two sets of accounts here, don't we?
   A.  Yes.
   Q.  We have the branch accounts?
   A.  Yes.
   Q.  And we have the MoneyGram accounts?
   A.  Yes.
   Q.  In the scenario where funds have been committed in the
       MoneyGram domain, what happens to the MoneyGram
       accounts?
   A.  So that would align with Post Office but there's
       back-end reconciliation not the front end.
   Q.  I'm sorry, but could you just focus on my question.
   A.  Sorry, okay.
   Q.  We've got the branch's accounts setting out what the
       branch's position is by reference to its transactions
       and then we've got MoneyGram's account setting out what
       their position is by reference to the transactions they
       are aware of.
   A.  Yes.
   Q.  Now, in the scenario that's described here, where funds
       have been committed in the MoneyGram domain, what would
       the MoneyGram accounts be showing?
   A.  If it has been committed they would be saying it has
       actually been transacted.
   Q.  Okay, so the MoneyGram accounts would show
       a transaction --
   A.  Yes.
   Q.  -- involving the branch.
   A.  Yes.
   Q.  But going back to the branch accounts, would the branch
       accounts show that transaction?
   A.  From reading this then no.
   Q.  Right.  So there would be a discrepancy, there would be
       a --
   A.  Yes.
   Q.  -- failure to reconcile, would there --
   A.  Yes.
   Q.  -- between the branch system, which means Post Office --
   A.  Yes.
   Q.  -- and MoneyGram, is that right?
   A.  That's right.
   Q.  Thank you.
           If we could then go on to {F/1502} please.  Do you
       remember Mr Green took you to this, do you remember
       seeing this yesterday?
   A.  I do, yes.
   Q.  If I could move on to page 29 of that document
       {F/1502/29} you will see it is headed "Latency
       resolution plan" and I recall Mr Green taking you
       straight down to the bottom entry, "Reconciliation",
       "Duplicate transactions are created ..." and as
       I understand it what he is saying is that's precisely
       the kind of duplicate transaction that may have happened
       in this particular case that we have been talking about.
       But if we look at the row headed "Reconciliation" it
       says:
           "Duplicate transactions are created in MG systems as
       a result of the Post Office timeouts."
   A.  Yes.
   Q.  Now, what does "MG systems" mean?
   A.  That's MoneyGram.
   Q.  So could you explain what's actually happening in this
       when these transactions are being duplicated?
   A.  So that's registered in MoneyGram as if they have been
       transacted.
   Q.  So in whose system is the duplicate transaction being
       created?
   A.  In MoneyGram's.
   Q.  Is there any description, any suggestion that there's
       a duplicate transaction being created in a branch's
       account?
   A.  Not from this, no.
   Q.  Thank you.
           Mr Green this morning mentioned two bugs,
       Dalmellington and Callendar Square, and that's in the
       context -- do you remember which witness this was in the
       context of?  I'm afraid I don't.
   A.  It was the transfers.  It was Latif.
   Q.  It is Mr Latif, I'm so sorry.  I'm being rather scatty.
           My learned friend put to you that you should have
       borne in mind -- he asked you whether you had and
       I think sought to suggest that you should have borne in
       mind two particular bugs, the Dalmellington bug and the
       Callendar Square bug.  I would like to ask you about the
       Dalmellington bug first.  Could I ask you to go to
       {F/1426}.  This is a document which I suspect you have
       never seen before, but -- why don't I ask you to read
       the document.  It is a KEL, it's a Fujitsu KEL relating
       to the Dalmellington bug.  Might you have seen this
       before?
   A.  Not in this format, no.
   Q.  Okay, well, then I will give you a few minutes to read
       it.
           (Pause).
   A.  Okay.
   Q.  Now, when you were asked about Dalmellington I think
       Mr Green suggested to you that this should have been in
       your mind as a possible explanation for what was
       experienced by Mr Latif and you answered him by saying
       "Well, no this related to an outreach branch".  Did
       Mr Latif's branch involve any outreach element?
   A.  No.
   Q.  So was this a bug that in fact could have affected
       Mr Latif's branch?
   A.  No, no.  So this refers to rem in and out process.  We
       were talking about -- Mr Latif was transfers in and out
       which is within branch, so this -- so my understanding
       of this is this is an outreach only issue, not other
       branches.
   Q.  So in your view does it have any relevance --
   A.  No.
   Q.  -- to the problem that was experienced by Mr Latif?
   A.  No, it has no relevance at all.
   Q.  And let's just talk about the symptoms as well.  Do you
       recall what the symptom is, the problem that was created
       by the Dalmellington bug, if we can call it that?
   A.  So this was when it was booked out from the core and
       been booked in by the outreach and there would be -- it
       was to do with a button press, they were able to enter
       the button again and that was doubling it.
   Q.  So what's happening is that we have money coming out
       from the main branch stock unit and going into the
       outreach stock unit.
   A.  Yes.
   Q.  And what you're suggesting is that the result of the
       Dalmellington bug was that too much would go into the
       stock unit --
   A.  Yes.
   Q.  -- the outreach stock unit, is that right?
   A.  Yes.
   Q.  In your view does that bear any similarity to the
       problem that Mr Latif says he experienced in relation to
       not being able to transfer money into --
   A.  No.
   Q.  -- the stock unit?
   A.  Mr Latif said that when he transferred from one to the
       other the transfer in wasn't displaying at all and
       therefore he reversed, is what he said, so this is very
       different to this.
   Q.  It's very different.  So in your view is the
       Dalmellington bug something that should have been borne
       actively in mind when you were addressing Mr Latif's
       data and what happened in his branch?
   A.  No.
   Q.  Thank you.
           Now, Callendar Square, it was also put to you that
       you should have borne Callendar Square in mind.  Are you
       familiar with the Callendar Square bug?
   A.  I have forgotten the details exactly of Callendar Square
       actually.
   Q.  Let me remind you.  If we could go to D2 tab 4 please,
       page 20 {D2/4/20}.  This is Mr Coyne's report and
       I would just like you to read paragraph 3.34 of his
       report please, Mrs van den Bogerd.
           (Pause).
   A.  Okay.
   Q.  You will see the year of the bug, it was in 2005 and
       fixed in March 2006.
   A.  Yes.
   Q.  Do you recall the year that Mr Latif suffered his
       problem?
   A.  Exactly no, but it was later than that.  Sorry, I would
       have to refer to my statement.
   Q.  Yes, why don't we.  Your statement doesn't actually --
       the months you look at are June, July and August 2015.
   A.  That's right, yes.  Sorry.
   Q.  So this is a bug that happened beforehand and Mr Coyne
       says that it occurred in Legacy Horizon and he refers to
       a data communication error in Riposte.  Now, when
       Mr Latif suffered his problem was he using
       Legacy Horizon?
   A.  No.
   Q.  And this may be beyond your expertise and if it is
       please tell me, but as far as you are aware does
       Horizon Online, the system that Mr Latif was using, use
       the Riposte system?
   A.  I don't think it does, but that would be beyond me.
   Q.  Very good.  But bearing in mind the year and the system
       on which this bug occurred, do you feel that you should
       have actively borne in mind the possibility that
       Mr Latif's problem was caused by the Callendar Square
       bug?
   A.  No.
   Q.  Thank you.
           Now, you were asked some questions about Mr Tank and
       in particular on a number of occasions you said Mr Tank
       should have kept the disconnected session receipt.
       Do you remember that?
   A.  Yes.
   Q.  Now, why -- I'm not criticising Mr Green here but why do
       you think that's significant?  What difference would it
       have made if Mr Tank had had the receipt?
   A.  Because that -- 1, it tells the postmaster what to do in
       that particular situation, whether to give cash to or
       take cash from, and that helps him identify then -- so
       for instance in this situation if the transaction is
       missing then he would be able to use that receipt to be
       able to identify that as well, so he's got much better
       position of what's happened in his branch.
   Q.  And with the benefit of that receipt, when a postmaster
       phones into the helpline with a problem with his
       accounts, does that assist the process --
   A.  Yes.
   Q.  -- the fact that there is a receipt?  Perhaps you could
       explain why that is so?
   A.  So ringing into the helpline to say that "I have an
       issue, I have a disconnected receipt, a recovery
       receipt", whatever the situation is, that's very
       different to saying "I have a discrepancy of £200" --
       195 in this instance -- "and I don't know what it is",
       it's a very different scenario and the helpline will
       respond differently to that.
   Q.  Very good, thank you.
           Now I would like to ask you about quite a lengthy
       correction regarding the Helen Rose report.  Do you
       remember the Helen Rose report?
   A.  Yes, I do, yes.
   Q.  The report is at {F/1082}.  If we could look at page 1
       of that report {F/1082/2}, at the bottom of the page you
       will see that -- well, from halfway down it is headed
       "Questions asked and extracts from various emails in
       response".  It is worth remembering that this was sent
       on 12 June and so what is contained below that
       heading -- perhaps I can ask you: what do you think --
       the text in blue -- the text in black and the text in
       blue, what do you think that represents?
   A.  That was the response from Gareth Jenkins to the -- so
       I think what Helen has done is just cut and pasted that
       into this document.
   Q.  I see.  So it would have been various emails that were
       written -- exchanges of emails that were written leading
       up to 12 June, is that right?
   A.  Yes, and you can see that she has referenced the email
       at the bottom there.  30 January 2013.
   Q.  So it's quite a long time, that's quite an early email,
       isn't it, it's six months before this report was
       written?
   A.  Yes.
   Q.  And if we look at that paragraph -- well, if we look at
       the previous paragraph, this is Mr Jenkins, he says
       about halfway down:
           "The fact that there is no indication of such
       a receipt in the events table suggests the counter may
       have been rebooted and so perhaps may have crashed in
       which case the clerk may not have been told exactly what
       to do."
           So there he appears to be suggesting that there was
       no receipt?
   A.  Yes.
   Q.  And it was put to you that there was no receipt and you
       resisted that suggestion?
   A.  Yes.
   Q.  But it was put to you on the basis of this document that
       there was no receipt.  Could we just go to the second
       page of this document -- it is the third page, isn't it,
       I see {F/1082/3}.  In the second paragraph down it says:
           "The files ... are part of the standard ARQ
       returned.  Rows 141 to 143 ... clearly show a reversal.
       Also row 70 ... shows that session 537803 ... has been
       recovered and this event has the same time stamp as the
       reversal session.  Also row 71 of events ... shows that
       a receipt was generated from the session 537805 (not
       explicitly, but it was the only session at that time)."
           So you didn't have an opportunity to look at this in
       the context of what was being suggested by my learned
       friend --
   A.  Yes.
   Q.  -- but now that you have seen this email, or the quote
       from this email which I apprehend is a later email, what
       would your reaction be to the suggestion that this
       document indicates that no receipt was actually
       produced?
   A.  That does say that receipt was produced.
   Q.  Thank you.
   MR JUSTICE FRASER:  Do you want to stop and revisit it at
       2 o'clock?
   MR DE GARR ROBINSON:  My Lord, yes please.
   MR JUSTICE FRASER:  I'm afraid you're going to have to come
       back.  We will stop until 2.
           Now, Mr De Garr Robinson, as you will have noted
       from my approach to Mr Green last week in terms of
       re-examination, it has to be kept within a narrow
       compass.  I know Mrs van den Bogerd is a very important
       witness so I'm not cutting you short on your
       re-examination, but just bear that in mind.
   MR DE GARR ROBINSON:  I will endeavour to ensure it is not
       lengthy, my Lord.
   MR JUSTICE FRASER:  2 o'clock.  Thank you very much.
   (1.00 pm)
                    (The luncheon adjournment)
   (2.00 pm)
   MR DE GARR ROBINSON:  Mrs van den Bogerd, we were looking at
       the Helen Rose report and I just have a few more
       questions about that.  The reference is {F/1082}.  If we
       could look at -- it may be the second page but it is the
       first page of the text {F/1082/2}.  At the bottom of the
       page you will see there are these three paragraphs which
       are pale blue which are copied and pasted from some
       earlier Gareth Jenkins email.  He says in the first
       paragraph that:
           "This shows that session 537803 was successfully
       saved to the BRDB, but when the user JAR001 logged on
       again recovery reversed the session in session 537805."
           And in relation to that it was put to you and you
       fairly accepted when you were being examined by Mr Green
       that it was Horizon that caused the reversal of the
       session.
           Could you explain what's meant by recovery reversed
       the session; could you explain what that means?  That's
       a slightly open question but I have to ask you open
       questions.
   A.  It means that the recovery process itself, what's
       triggered as a result of the loss of connectivity here
       then actually reversed that transaction and that was
       triggered by the postmaster logging on after the system
       had come back up.
   Q.  So it was the postmaster logging on --
   A.  Yes.
   Q.  -- that was the start of the process but it was then
       recovery which when faced with the postmaster logging on
       caused the transaction to be reversed, is that right?
   A.  That's right.
   Q.  Then in the third paragraph down:
           "The reversal was due to recovery ... so this was
       not a explicit reversal by the clerk."
           It may be that you don't, but do you know what
       Mr Jenkins means when he says "explicit reversal by the
       clerk"?
   A.  What he means is that the, it would be the postmaster in
       this instance, didn't actually go in and do a reversal
       himself, so he didn't actually go and look for the
       session number to do the reversal.
   Q.  I see.  Now, when you were asked some questions about
       this document you referred to an email.
   A.  Yes.
   Q.  And it might be that I know the email you were referring
       to.  Could I ask you to go please to {F/1095.1}.  I will
       just give you a minute or two just to familiarise
       yourself with that document.
   A.  That's right, yes.
           (Pause).
           Yes, I do recall the email.
   Q.  Right.  If one goes halfway down this page you will see
       it is an email chain but there's an email there from
       John Armstrong to Ron Warmington.  Can you tell me
       John Armstrong was?
   A.  He is the postmaster at Lepton.
   Q.  And who is Ron Warmington?
   A.  He is one of the directors at Second Sight.
   Q.  If we could go now to page 6 of that document
       {F/1095.1/6}, this is an email from Mr Warmington to
       Mr Armstrong and it is dated 25 June, so this is nearly
       two weeks after the Helen Rose report, just so we can
       locate ourselves in the chronology, and Mr Warmington
       from Second Sight says:
           "John, as I mentioned we are including your incident
       in our interim report to be delivered to MPs on July 8.
       It's one of only four spot reviews that we are at this
       stage disclosing."
           Can you explain what a spot review is?
   A.  It was terminology that Second Sight used for a deep
       dive into an issue that had been raised, so they
       investigated an issue and they termed it a spot review.
   Q.  I see.  And if we go on to the next paragraph:
           "Here is what we have in the report on your
       incident.  If you see anything at all here that needs to
       be corrected, or could be improved in any way, please
       let me know and I'll change it."
           Then there's some text I need not take you to and
       then going down to below the next paragraph he says
       "Report extract follows here" and if you could read the
       bottom of that page and then the top of page 7
       {F/1095.1/7}.  You can tell us when you get to the
       bottom of the page.
   A.  Under "SR001", from there, is that where you mean?
   Q.  And then when you get to the bottom tell us and we will
       go on to page 7.
   A.  Yes, turn the page please.
           (Pause).
           Okay so the bottom of that paragraph, yes?
   Q.  Okay.  And then we have "SR001 summary of POL's
       response" and then we have:
           "POL's 10-page response asserts that the Spot
       Review:
           "Does not demonstrate any failing in Horizon.

           "Principally asks whether a SPMR will be properly
       notified about automatic reversals of transactions when
       Horizon is unable to connect to the data Centre."
           And then I don't need to take you to the next
       paragraph but in the next paragraph but one says:
           "POL further asserts in its response that the root
       cause of the difficulties suffered by the SPMR was his
       failure to follow the on screen and printed instructions
       given by Horizon.  POL claimed to be confident that the
       SPMR knew that some transactions had been automatically
       reversed because ..."
           First:
           "The branch had been suffering connectivity issues
       in the run-up to the incident in question ..."
           Do you see that?
   A.  Yes.
   Q.  And then if we go over the page {F/1095.1/8}, second:
           "When the transactions in question first failed to
       be processed (because Horizon could not get a response
       from the data centre), Horizon asked the SPMR whether he
       wished to cancel or retry the transactions, in response
       to which the SPMR opted to retry the transactions ..."
           Next:
           "When the transactions failed again, the SPMR opted
       to cancel the transactions.
           "Horizon then automatically disconnected and printed
       a 'disconnect' receipt that showed the transactions that
       had been automatically reversed."
           Next:
           "A standard customer receipt was not produced and
       POL asserts that this should have told the SPMR that the
       full transaction had not proceeded."
           And:
           "Following the disconnect the SPMR was required to
       log back on to Horizon and duly did so."
           And:
           "Following the log on, and as part of the standard
       recovery process, Horizon printed a 'recovery' receipt
       which again showed the transactions that had been
       reversed and those that had been recovered."
           Now, does that refresh your memory as to why you
       said you had separate knowledge as to whether or not
       Mr Armstrong received a ..?
   A.  Yes.
   Q.  I see.  Let's move back to page 6 {F/1095.1/6}.  This is
       the beginning of the same email we looked at and we read
       the first big paragraph of the email.  I would like to
       read the second paragraph now:
           "Ian and I ..."
           I take it that's Mr Henderson?
   A.  That's correct.
   Q.  "... are particularly interested to know whether you
       received the printed disconnect and recovery 'receipts'
       that POL refers to in its response ..."
           So there's the question that Mr Warmington asks
       Mr Armstrong and Mr Armstrong's reply is at page 4
       {F/1095.1/4} and can I ask you please,
       Mrs van den Bogerd, just to read the email of Tuesday
       25 June from the middle of the page downwards and then
       when you've got to the bottom tell us and I will ask you
       to read half of the next page.
           (Pause).
   A.  Yes, I have read that.
           (Pause).
           Yes.
   Q.  Let me know when you are done?
   A.  Sorry, I'm done.
   Q.  You're done?
   A.  Yes.
   Q.  Now, first of all -- I hope this is uncontroversial --
       what Mr Armstrong seems to be saying there is that he
       did receive some disconnected session receipts.
   A.  Yes.
   Q.  What would the fact of receiving these disconnected
       session receipts have told him?
   A.  It would have told him that the BT bill hadn't actually
       been processed.
   Q.  So we now have more information than Mr Jenkins had in
       the emails that are attached or rather quoted in the
       Helen Rose report.  If we could go to page 3 of this
       document please {F/1095.1/3}, this is the second page of
       an email from Mr Baker from the Post Office,
       Simon Baker.  Do you know who Simon Baker is?
   A.  He was a project manager at the time.
   Q.  I see.  You will see at the top of the page it says
       "Regards, Simon" so it is the end of his email but we
       have Gareth's comments below, that's Mr Jenkins'
       comments.  There's a reference to GMT and British Summer
       Time, this is because the logs used GMT even though BST
       is in fact in force.
           Could I ask you please to read paragraphs 1 through
       to 9.
           (Pause).
   A.  Yes, I have finished.
   Q.  You will see at the bottom in paragraph 9 Mr Jenkins
       says:
           "Therefore a message asking for a retry or cancel
       would have been shown at about 09.36.25.  At this point
       'cancel' must have been selected thus producing the
       disconnected session receipt.  This fits in with the
       time of 10.36 in the email below."
           Would you care to comment?  Does that accord with
       your understanding, or do you have a different
       understanding?
   A.  No, that's my understanding.
   Q.  That's your understanding?
   A.  Yes.
   Q.  So having gone through those documents can I ask you now
       to go back to your witness statement please to
       paragraph 154 {E2/5/35} and this is where you say -- I'm
       going to ask you what you meant by this now.  You say in
       paragraph 154:
           "At paragraph 5.175 of his report, Mr Coyne has
       referred to a report prepared by Helen Rose ... in the
       context of failed reversals.  The extracts taken from
       the report by Helen Rose referred to by Mr Coyne are
       taken out of context and mistakenly claim that the
       relevant reversal was issued in error by Horizon not the
       subpostmaster."
           I just want to give you an opportunity -- Mr Green
       was pressing you to accept that actually what you had
       said in that sentence was wrong.  I would like to give
       you an opportunity to explain to the court what you
       actually meant.
   A.  What I meant was that the actual reversal was part of
       that recovery and it had actually taken place as it
       should have taken place, which is what I meant in that.
       So it wasn't a failed reversal because it actually had
       happened as it should have happened, but I accepted in
       there that the -- it wasn't obvious to the postmaster at
       the time that what had happened -- that he hadn't --
       because it didn't show that he had actually -- it showed
       that he had done it and he knew he hadn't done what we
       referred to earlier was an explicit reversal.  That's
       what I meant in that.
   Q.  I won't take any more time up on that question.
           Just two more small matters, Mrs van den Bogerd.
       First of all, it was put to you by Mr Green on page 80
       of yesterday's transcript {Day5/80:16} that Mrs Mather's
       statement shows that Credence contains a log of every
       key that's pressed --
   A.  Yes.
   Q.  -- on the counter and I would like to take you to
       Mrs Mather's statement, if I may.  That's E2, tab 8 and
       I would like to go to page 2 of that statement please,
       paragraph 9 {E2/8/2}.  You will see what Mrs Mather
       actually says in the first sentence.
   A.  Yes.
   Q.  Now, the reason why I'm putting this to you is because
       Mr Green put to you that Credence shows every single
       keystroke that's pressed by the postmaster and you
       accepted what he said and he put it to you on the basis
       that this is what Mrs Mather said.  Could I ask you this
       question: is it the case that Credence gives an account
       of every single key that's pressed by the postmaster in
       branch?
   A.  That's not my understanding, not every single stroke.
   Q.  I'm grateful.
           Then one small final question.  It is based upon the
       document at {F/555}.  You were taken to this document by
       Mr Green and it was suggested to you that Horizon Online
       was about reducing operating costs.  Do you remember
       that?
   A.  Yes.
   Q.  I would like to explore that very briefly with you.
       Could we go to page 6 of that document {F/555/6}.  It is
       "Purpose of the HNG programme":
           "To deliver a significant reduction in the total
       annual cost of ownership of Horizon, whilst ensuring the
       system remains fit for purpose in the 21st Century ...
       in a dynamic and changing environment."
           If we go down to the second half of the page:
           "Recognising that HNG is not operating in a silo -
       life goes on in the business - initiatives like Rural
       Strategy, Network Strategy, Packet Strategy, Re-org all
       have potential to bring around change ... to the
       programme.
           "HNG should also enable the business to reduce the
       time to market for major developments; should be
       a simpler system that is easier to change and service
       manage; should be capable of supporting segments and
       zones.
           "The programme is about developing the current
       Horizon system to meet business drivers, and is
       a balancing act between reducing cost and building
       a system fit for our needs we will cover this in
       a little more detail in next few slides."
           What I'm going to be asking you, Mrs van den Bogerd,
       is whether the introduction of Horizon Online was just
       about saving cost or whether it was about other things
       as well, but perhaps I can ask you that question now.
   A.  Yes.
   Q.  Having read that page --
   A.  It wasn't just about saving costs.  There were other
       benefits for us too as an organisation.
   Q.  Could you tell the court what those other benefits or
       some of those other benefits were?
   A.  So 1, it took it to an online system, where it hadn't
       previously been and it did enable us to make changes
       a bit more quicker to the system itself.  I think we did
       look at some screen changes as well, I think.  But it
       was really about a better system going forward, that's
       what it was about.  So there were clear advantages to
       being able to get changes through the system and get
       products changes done more quickly.
   Q.  If we go to page 8 {F/555/8} -- well, my Lord, I see the
       time.  The document says what it says.  I do believe
       I have no further questions for this witness.
   MR JUSTICE FRASER:  All right.
                 Questions from MR JUSTICE FRASER
   MR JUSTICE FRASER:  I just have a couple of questions.
           A couple of times this morning when you were being
       asked about the detail of your witness statement and
       incorrect references to various documents, et cetera,
       you said you didn't have a lot of time or you didn't
       have a huge -- one of your exact answers was that you
       didn't have a huge amount of time to put this together.
   A.  Yes.
   MR JUSTICE FRASER:  Was there a time limit on you preparing
       your witness statement?
   A.  In as much as getting the information to investigate the
       issues, that was what I meant.
   MR JUSTICE FRASER:  And what sort of time limit or pressure
       was that?
   A.  To pull the information and to be able to look into the
       detail of what the claimants' issues were.  It was that.
   MR JUSTICE FRASER:  And is there anything else you would
       like to tell me about that, about time pressure?
   A.  It's just it was -- getting the team to put together the
       information.  So because there were a number of issues
       that needed to be investigated, which was why I needed
       to use the team -- I would ordinarily have used the team
       anyway to do that -- it was then me being able to get
       into the detail to be able to meet the court deadline to
       get the witness statement submitted.
   MR DE GARR ROBINSON:  Does your Lordship want to ask this
       witness about deadlines for producing the evidence?
   MR JUSTICE FRASER:  No, not particularly.  I'm going to give
       you an open opportunity to ask any questions arising out
       of my questions.
           Another phrase that pops up in different documents
       and you may not necessarily know what it is, but I just
       thought you were the best person to ask.  What's
       "operational risk"?  What does that mean when the
       Post Office uses it in internal documents?
   A.  It means something -- operational risk would mean
       perhaps losing power across the whole of the Horizon
       system, so if we did have a failure in terms of
       connectivity, that would cause us operational risk.
       It's about us being able to maintain service, so key
       systems going down, for instance, would not allow us to
       be able to operate: the back office systems, POLSAP or
       Horizon clearly for the front offices as well.
   MR JUSTICE FRASER:  So it is a risk to the way the
       Post Office business is operated --
   A.  Yes.
   MR JUSTICE FRASER:  -- is that how I should read it?
   A.  Yes.
   MR JUSTICE FRASER:  You mentioned in that first answer
       losing power across the network.  Losing power to an
       individual branch, or an interruption of power at an
       individual branch, would that be an operational risk or
       not?
   A.  Yes, it would be.
   MR JUSTICE FRASER:  That would be?
   A.  Yes, we would class that as operational risk as well.
   MR JUSTICE FRASER:  Right.  And then you were asked about
       the list of people who had helped you with your witness
       statement and you said I think there were about nine or
       ten and then you gave us a list from memory.  Was there
       any reason that you didn't list those people in your
       witness statement?
   A.  I didn't think it was necessary.
   MR JUSTICE FRASER:  All right.
           Then can we call up a document please, {F/1449},
       which is a document we looked at yesterday.  Do you
       remember Mr Green asked you some questions about this
       document?
   A.  Yes.
   MR JUSTICE FRASER:  And it shows two different approaches
       depending on whether the amount of the miskey was
       greater than £150 or not?
   A.  Yes.
   MR JUSTICE FRASER:  Now, I'm a bit confused about the
       document which is just why I wanted to ask you.  Let's
       say the miskey is £200, am I right you would go down the
       right-hand branch as we look at it and the branch would
       be told that if the discrepancy, which in my example is
       £200, was still apparent when they were balancing they
       must make the amount good?
   A.  That's what it says on here.
   MR JUSTICE FRASER:  And I recall from the common issues
       trial that means they would, for example, have to pay
       the £200 in, is that right?
   A.  Yes.
   MR JUSTICE FRASER:  Right.  If we look at the left branch,
       which is for less than 150, so let's say that that's
       £50, there seems to be a little bit more to it, which
       includes the possibility of a transaction correction.
   A.  Yes.
   MR JUSTICE FRASER:  Have I read that correctly?
   A.  Yes.  This is a bit confusing on -- I haven't seen this
       before.  Because to me the significance of 150 is the
       amount that you can settle centrally or not and
       therefore in reading it that way, which is how I would
       interpret that -- so if you said it was over 150, what
       should be in that box is when it says "Must make good"
       the option in there should be to settle centrally
       because it's over the 150.
   MR JUSTICE FRASER:  Well, let's say we read it in that way,
       so that includes settling centrally.
   A.  Yes.
   MR JUSTICE FRASER:  Go back to the left-hand box, that seems
       to include within it the possibility of a transaction
       correction being sent out later, because it says that in
       the last four lines.
   A.  Yes.
   MR JUSTICE FRASER:  That entry is not on the right-hand
       side, so am I correct in reading this chart that if it
       is above 150, that option is not potentially available,
       but if it's below 150, a transaction correction might be
       sent out in the future?
   A.  It might be sent out if it is over 150 as well.
   MR JUSTICE FRASER:  Well, that then takes me to my next
       question: but it doesn't say that on the right-hand
       side.
   A.  No.
   MR JUSTICE FRASER:  Can you think of any reason why that
       might not be in the box on the right-hand side?
   A.  The only reason I can think of that is that this is an
       incomplete document and, as I say, I've never -- I can't
       recall ever seeing this, but that wouldn't make sense to
       me reading that as currently presented.
   MR JUSTICE FRASER:  All right.  Thank you very much.  That's
       very useful.
           Now, any questions arising out of that?
   MR DE GARR ROBINSON:  Just about timing.
          Further re-examination by MR DE GARR ROBINSON
   MR DE GARR ROBINSON:  Mrs van den Bogerd, when you made your
       witness statement do you recall that it was in response
       to some witness statements made by the claimants?
   A.  Yes.
   Q.  And were those witness statements expected?
   A.  No.
   Q.  Did you expect to be dealing with that kind of evidence
       when it came?
   A.  No.  No, I didn't at the time.
   Q.  So these witness statements came in.  Do you recall how
       much time you had to prepare the witness statement in
       response?
   A.  I don't remember exactly but it would have been from
       memory about three weeks or so.  I can't remember
       exactly.
   Q.  And you say three weeks, but before you could properly
       start work on that responsive evidence did some steps
       have to be taken to --
   A.  Yes.
   Q.  -- obtain the information?
   A.  Yes, we needed to get the information, the transaction
       data and NBSC logs, everything that we would normally
       get.
   Q.  And do you recall how long that took?
   A.  Well, transaction -- if we get our files they can take
       about a fortnight to come, just to get -- so it would
       take --
   Q.  I see.  Is your evidence then that after that time had
       been taken up you then had about three weeks to actually
       prepare a witness statement?
   A.  I don't even think it was that long actually, if
       I remember.  So ordinarily we would get the information
       and then we would have time to do that properly.  This
       was quite pressurised because of those statements coming
       in unexpectedly and then having to get that, so it was
       under pressure.
   Q.  Thank you very much.
   MR JUSTICE FRASER:  Thank you very much.  You are now free
       both to leave the witness box and obviously to discuss
       the case again if you want.
   A.  Thank you.
   MR JUSTICE FRASER:  Right, Mr De Garr Robinson.
   MR DE GARR ROBINSON:  My Lord, I call Ms Dawn Phillips.
                   MS DAWN PHILLIPS (affirmed)
   MR JUSTICE FRASER:  Would you like to have a seat.
   A.  Thank you.
           Examination-in-chief by MR DE GARR ROBINSON
   MR DE GARR ROBINSON:  Ms Phillips, there should be a bundle
       of documents to hand.  If I can ask you to go to tab 3
       of that document {E2/3/1}, I believe the first page of
       that document should describe itself as a witness
       statement of Dawn Louise Phillips, do you see that?
   A.  Yes.
   Q.  Is that your name and address there at the front?
   A.  Yes, it is.
   Q.  If you go to page 3 of the statement, is that your
       signature at the end?
   A.  It is.
   Q.  And is this witness statement true to the best of your
       knowledge, recollection and belief?
   A.  Yes, it is.
   Q.  Ms Donnelly will have some questions for you.  Thank you
       very much.
                 Cross-examination by MS DONNELLY
   MS DONNELLY:  Good afternoon, is it Ms or Mrs Phillips?
   A.  Ms.
   Q.  Ms Phillips, in your witness statement you say you are
       team leader for agent accounting and Santander banking?
   A.  Yes.
   Q.  So I will deal with those separately if I may, taking
       agent accounting first.  Can I understand from your
       paragraph 4 {E2/3/2}, you say:
           "... I oversee the process of recovering the losses
       that postmasters have declared in branches."
           Are you the team leader for agent accounting?
   A.  Yes.
   Q.  And when you are referring to "recovering the losses",
       are you referring there to the amount that they have
       settled centrally?
   A.  Yes.
   Q.  If you look at paragraph 7 of your witness statement
       which is at {E2/3/2}, do you see three lines down there
       you refer to "chosen to settle shortfalls centrally"?
   A.  Why he.
   Q.  You fairly recognise, don't you, at paragraph 9 that
       there's no option for the postmaster to dispute the
       discrepancies on Horizon, or record that they have
       raised a dispute on Horizon?
   A.  They chose to settle centrally rather than make good is
       what I meant.
   Q.  Yes, settle centrally rather than put the cash in
       straight away?
   A.  Yes.
   Q.  And that system remains the case today -- your witness
       statement is September.
   A.  Yes.
   Q.  And that remains the system.
           Where you describe the agent accounting team, in the
       common issues trial back in November of last year we
       heard reference to the current agent debt team --
   A.  That's the old name.
   Q.  Does the substance of what you do remain the same?
   A.  Yes.
   Q.  So there's really been a semantic change from "debt" to
       "accounting"?
   A.  (Nods).
   Q.  And has the same thing happened with former agent debt?
   A.  Yes, it's former agent accounting.
   Q.  But what that team is doing is debt recovery from
       subpostmasters who have left or been terminated?
   A.  Not just debt recovery, there are gains settled
       centrally as well that we also deal with.
   Q.  The real substance of your statement is the dispute
       process as it now is.
   A.  Yes.
   Q.  And looking at your paragraphs 7 and 8 {E2/3/2} we see
       you refer the court to two letters, one at the end of
       paragraph 7 and one at the end of paragraph 8.1.  Now,
       those versions that you are referring to there, do you
       know when they were introduced?
   A.  1 September 2014.
   Q.  And are they still the current versions?
   A.  Yes.
   Q.  So if we could look please at { F/1832}.  Do we see
       there the first of the letters that would be sent?
   A.  Yes.
   Q.  Just from looking at the first four paragraphs we see:
           "Please find the attached statement that confirms
       the amount you owe to Post Office Limited.  The
       statement provides a breakdown of how you have incurred
       the total owed amount.
           "There are a number of ways you can repay the money
       to us; you will find all the details of the different
       methods of payments on the reverse of this letter.
           "If you would like to repay the money by
       debit/credit card ..."
           Then it goes on:
           "We look forward to hearing from you within the next
       7 days to arrange payment."
           The wording used in this letter is as if the amount
       is a debt to Post Office that needs to be repaid?
   A.  Yes.
   Q.  And the last paragraph says:
           "If you have any queries regarding the content of
       this letter, please contact one of my team or drop us
       an email at agents.accounting.team@postoffice.co.uk and
       we will do our best to help you."
           Now, what's missing from that letter is any mention
       of a dispute process there?
   A.  Yes.
   Q.  If you look at the first paragraph again, "Please find
       the attached statement", that's not included with the
       letter that you have exhibited which is this version,
       but it says it provides a breakdown, so what information
       would be shown on that statement?
   A.  It would be each individual amount settled centrally in
       the previous trading period, so it could be one branch
       discrepancy settled at the end of the month, several
       transaction corrections and/or an invoice.
   Q.  So if we go to {F/1257.2}, we see here an example
       I think of the same letter that we were just looking at
       which was sent to Mr Tank.  If we go to page 3
       {F/1257.2/3}, is that the statement, the form of the
       statement that would be attached?
   A.  Yes.
   Q.  So here there has been one amount settled centrally and
       it appears like this, "branch discrepancy" and then the
       date and the amount.
   A.  Yes.
   Q.  Again, this statement doesn't have any indication that
       it is possible to dispute this amount, does it?
   A.  No.
   Q.  Going back to your witness statement at {E2/3/2}, you
       say at paragraph 8.1:
           "If we do not receive a response within 7 days we
       send a second letter."
   A.  Yes.
   Q.  And as I understand from your witness statement there
       aren't any phone -- later on you refer to some phone
       calls might be made, but there aren't any phone calls
       in-between this stage, are there?
   A.  No.
   Q.  And when you say "if we do not receive a response", is
       that your team?
   A.  Yes.
   Q.  If we go then to {F/1833}, this is the second letter
       that you have exhibited.
   A.  Yes.
   Q.  "We wrote to you previously requesting payment of the
       amount you owe to Post Office Limited.
           "We are concerned that you have not contacted us to
       arrange payment.
           "Payment is now overview due and needs to be made
       within 7 days of this letter.
           "As previously mentioned, there are a number of ways
       you can repay the money to us.  These are listed on the
       reverse of this letter."
           So we have all the same language again, haven't we,
       that this is a debt?
   A.  Mm-hm.
   Q.  Then the next paragraph:
           "We have to advise you that if we do not receive
       payment or your instructions within this timescale, this
       may be escalated to your contracts advisor for further
       action."
           So what's that supposed to mean?
   A.  If they don't make a payment or raise a dispute with us
       then we could send it to the contracts advisor for
       further action.
   Q.  And in your answer you have referred to "raise
       a dispute", but again there's no mention of dispute, is
       there, in this letter?
   A.  Well, in both letters it mentions that if they have any
       enquiries they can contact us.
   Q.  It's a bit different, isn't it, to say "If you have any
       queries regarding the content of this letter" -- it's
       a bit different writing that to "If you want to raise
       a dispute about this amount then please find enclosed
       the dispute process"?
   A.  Yes.
   Q.  Are these two letters usually successful in getting
       money in?
   A.  Sometimes it is resolved even before the second letter
       goes out.  It varies.
   Q.  But by the time this second letter is sent, normally the
       subpostmaster has paid?
   A.  Yes.
   Q.  If we look at your witness statement again, so {E2/3/2},
       paragraph 9, in the second sentence there you say that
       postmasters are able to call the NBSC helpline or your
       team and raise a dispute.
   A.  Yes.
   Q.  The letters we just looked at don't refer to calling the
       helpline, do they?
   A.  No.
   Q.  But it is possible, isn't it, that a subpostmaster might
       have raised a dispute on the helpline before you send
       your first letter?
   A.  If they have raised a dispute on the helpline before we
       have sent our letter then it is before they have
       actually settled it centrally.
   Q.  Well, you don't send the letter the second they settle
       centrally, do you?
   A.  We send it on the Monday after, so unless they actually
       made the error on the last day of their trading period
       then it's likely that they made the error in the
       previous few weeks.
   Q.  So they could have identified that they've got an
       unexplained discrepancy say two weeks before the end of
       their trading period.
   A.  Yes.
   Q.  And they call the helpline at that point and say "I want
       to dispute this, this is going to affect my accounts,
       I'm not responsible for this, it is a system error" and
       two weeks later, nothing has been resolved, they have to
       settle centrally, you then send them the letter on the
       Monday?
   A.  Yes.
   Q.  So what you don't do is go and check the helpline logs
       to see if they have actually already raised a dispute?
   A.  We do actually, every Friday we run a report to see who
       has settled something centrally and although we can't
       check every amount we do try and check the majority of
       the larger ones.
   Q.  Sorry, I think you answered by reference to whether you
       check who has settled centrally --
   A.  Yes.
   Q.  -- and I was asking whether you routinely check the
       helpline logs?
   A.  Yes, I do.  Yes we do, as a team.
   Q.  For the larger ones you check the helpline logs?
   A.  Yes.  Unfortunately there are too many to check every
       single one, but we do check the larger amounts.
   Q.  So your team has access to the helpline logs?
   A.  Yes.
   Q.  As I understand it, it depends on somebody in your team
       going to check the helpline logs rather than there being
       an automatic process from the helpline to notify your
       team?
   A.  If the branch have already settled centrally and tell
       NBSC that they want to dispute that, then they would let
       us know.
   Q.  But I was speaking about if a subpostmaster was raising
       a dispute in advance of settling centrally?
   A.  Then no, until he has actually settled it centrally then
       I wouldn't know about it.
   Q.  So subpostmaster could call up the helpline and say "I'm
       going to have a problem, I know this is going to affect
       my accounts, I'm not responsible for this amount, I want
       to dispute it" and then you still end up sending your
       letter out on the Monday following the settled
       centrally?
   A.  I can't speak for any calls that have happened like
       that.  I've never seen one.
   Q.  Well, you also don't look, do you?
   A.  Well, we do regularly check.  We check every week.
   Q.  But aren't you looking for when they have settled
       centrally calls?
   A.  Yes.
   Q.  You're not looking back and saying "Has there been any
       calls before that"?
   A.  They don't make -- we look for any call in the previous
       month that could relate to the loss that they have
       settled.  We don't specifically search for settled
       centrally calls.
   Q.  And so this is an ad hoc process, is it?
   A.  No, it's a weekly standard process.
   Q.  But which ones you decide to look at, or is there
       a specific amount?
   A.  We try and do everything above 5,000.
   Q.  Above 5,000?
   A.  Mm-hm.
   Q.  Right.  So you wouldn't bother checking the helpline log
       for £2,000?
   A.  It's not a case of not bothering, it's a case of how
       many have settled centrally and how many are large.
   MR JUSTICE FRASER:  I wonder if you could start again with
       the sequence, because I'm slightly puzzled.
   MS DONNELLY:  My Lord, I do apologise: the sequence from if
       a subpostmaster has raised a dispute prior to settling
       centrally?
   MR JUSTICE FRASER:  And what's checked and when.
   MS DONNELLY:  So if mid-way through trading period
       a subpostmaster contacts the helpline and says "I have
       a £2,000 discrepancy in my accounts that I know if it is
       not resolved is going to have to be settled centrally
       and I want to dispute that now", as I understand your
       evidence, after they in fact settle centrally, on the
       Monday morning your team will send them letter 1 because
       you wouldn't check the helpline logs for an amount less
       than £5,000?
   A.  I'm not saying it's always £5,000.
   Q.  But that's generally what would happen?
   A.  Yes.
   MR JUSTICE FRASER:  So if it is not always £5,000 what other
       figure limit would it be?
   A.  It depends.  We can sometimes check them all.  It
       depends how many fresh items have been settled centrally
       in that week.
   MR JUSTICE FRASER:  And do you get a list of these?
   A.  I can pull them from our finance system.
   MR JUSTICE FRASER:  Filtered by an amount that you decide
       that day?
   A.  No, we pull everything that's been settled that month.
   MS DONNELLY:  Do you get any transcripts of those helpline
       calls?
   A.  No, we have to search for them.
   Q.  There will be a call log, but those helpline calls are
       also recorded, aren't they?
   A.  Yes.
   Q.  Do you listen to the recording, or do you just look at
       how it's --
   A.  We just look at the log on the screen, not a recording.
   Q.  So the completeness and accuracy of what appears on the
       helpline log would be very important to your team?
   A.  Yes.  But if we were unsure we would pick up the phone
       and speak to the branch.
   Q.  Do you send to the subpostmaster the copy of the
       helpline log that you are looking at so they know what
       you've got?
   A.  No.  They made the call, we would assume they would know
       what it was.
   Q.  Well, what the subpostmaster doesn't know is how it has
       been recorded by the helpline?
   A.  Well, we'd discuss that -- if there was an issue and we
       had to call them we would discuss that with them.
   Q.  Is there any auditing of the process that you have just
       described being carried out by your team?
   A.  Yes, we have controls in place that we have to report
       and measure every month.
   Q.  So what are they?
   A.  So we've got an external firm called Traction that we
       have to report to and attach evidence that we have
       followed that process every week or month, whatever the
       actual process is supposed to be done to.
   Q.  So do you report -- sorry, did you say it is an external
       company?
   A.  PwC I do -- I can't remember the name of it but we have
       an external company that we have to load all our
       evidence onto.
   Q.  Do you report to them which value you have decided to
       look at the helpline logs for --
   A.  No.
   Q.  -- that week?
   A.  No.
   Q.  So that detail isn't recorded?
   A.  No.
   Q.  If we just look at paragraph 10 of your witness
       statement which is {E2/3/2}, you say there that:
           "If a postmaster calls my team after receiving our
       request for payment and explains they are raising
       a dispute, we will place a block on their account on the
       system until the dispute is resolved.  If they do this
       within 7 days we will not send the second letter
       referred to in 8.1 above, but a postmaster can dispute
       a shortfall at any stage in the process (not just after
       the first letter)."
           And then you go on at paragraph 11:
           "When a postmaster disputes a shortfall with my team
       we gather as much information as possible from the
       branch about why they are disputing it.  We also send
       them a branch dispute form to complete ..."
           Just pausing there, is that right, that you gather
       as much information as possible and also send them
       a branch dispute form?
   A.  Not necessarily.  If it is something like a miskey which
       is just a simple case of the branch recording it with
       helpline, we check with helpline that it has been
       received and gone to the appropriate team, no need for
       a dispute form, that's a simple one.  We only send the
       dispute forms when it is more in-depth and we have to
       pass the information on for someone else to investigate,
       so it is better coming from the postmaster in their own
       words, so that's when the form is sent.
   Q.  So say it is a more complicated one when you are going
       to require a branch dispute form.
   A.  Yes.
   Q.  Do you first get all the information from them and then
       send a branch dispute form for them to provide it all
       again?
   A.  No.  It would probably start it with a discussion on the
       phone.  If it looked like it was going to be
       a complicated one we would stop it there and say "We're
       going to send you a form out", we put a block on the
       debt immediately so all the process was halted and then
       send them a form and let them fill it in and send it
       back.
   Q.  So your paragraph 11 isn't quite right, if it's the
       branch dispute form --
   A.  No, we don't do both.  It's not required for all of
       them.
   Q.  You don't necessarily do both and you don't always send
       a branch dispute form?
   A.  Only if it is difficult.
   Q.  Since it was introduced -- I think you later tell us it
       was introduced at the beginning of 2018?
   A.  Yes.
   Q.  How many have you sent?
   A.  Sorry, I don't have that number with me.
   Q.  Broadly?
   A.  I wouldn't like to guess.
   MR JUSTICE FRASER:  Well, thousands, two?
   A.  No, probably at least five a month.
   MR JUSTICE FRASER:  Five a month?
   A.  Yes.  But bear in mind they are not just the dispute --
       they can just send us an email, it doesn't have to be
       the dispute form -- or a letter.
   MS DONNELLY:  Because your witness statement reads as if
       this is some sea change in approach.  If you look, for
       example, at paragraph 12 {E2/3/3}:
           "The branch dispute form was introduced to capture
       the key areas of information we would collect from
       a branch about a dispute.  Whilst the branch dispute
       form was only recently introduced at the beginning of
       2018, the process of collecting this key information
       from the branch has been in place
       since November 2016 ..."
           And it goes on.  It reads as if this was something
       new and improved and was happening all the time but --
   A.  It is improved.
   Q.  -- it's actually quite rare that you send it?
   A.  No, it is improved.
   MR JUSTICE FRASER:  Can you not overspeak each other.  It is
       both of you I think.  You put your question and then you
       give your answer.
   MS DONNELLY:  I do apologise, my Lord and Ms Phillips.
           It is actually quite rare, five a month or so, that
       you actually send out the branch dispute form, it's not
       quite the impression that you're giving in your witness
       statement, is it?
   A.  No, maybe not, but it still helps both us and the
       branch.  It has got simple things on there like who are
       we speaking to at the branch, because we have absentee
       postmasters but their name is on the account.  If it is
       somebody else running the branch we need to know who it
       is and their name and who to speak to.
   Q.  And without the branch dispute form you might not get
       that right?
   A.  Well, we send letters to the branch to the postmaster.
   Q.  Is there a financial limit that applies to your process
       of sending out the branch dispute form?
   A.  No.
   Q.  You are describing it as essentially something positive
       for subpostmasters, but if you look at paragraph 11 at
       the top of {E2/3/3}:
           "Branches are given seven days to return the
       completed form, otherwise we unblock the shortfall and
       send them another letter requesting payment."
           I mean in some ways a subpostmaster who receives
       a branch dispute form is in a worse place than if they
       didn't because now they're up against the clock to
       complete it in seven days or you unblock, is that right?
   A.  If they didn't send a dispute at all it would never be
       blocked and they would be up against the clock anyway as
       you put it.
   Q.  No, but in order for them to be sent a branch dispute
       form they have to dispute with your team in the first
       place?
   A.  Yes.
   Q.  So that's happened.  Then some, about five a month, get
       a branch dispute form sent to them and those ones now
       have to return it within seven days otherwise the amount
       is unblocked, even though they have already told you
       there's a dispute.
   A.  Again, it's seven extra days.  If they rang up and
       described the dispute to us there and then it would be
       blocked immediately and dispute -- and the dispute would
       take place, but if they can't explain it to us then they
       need to send the form in.
   Q.  Could we have a look at the form.  It is at {F/1831}.
       So who decided to introduce the form?
   A.  It was a joint between my team, my senior manager and
       the contracts advisor team.
   Q.  Who is your senior ..?
   A.  Alison Bolsover.
   Q.  Do you know was any part of this branch support
       programme related at all to what had gone before?
   A.  It came in a discussion between myself, Alison and the
       contracts advisors that we should all be working from
       the same book basically, so they use the form as well.
   Q.  If we have a look under "Dispute details", it says:
           "Please provide details of your challenge, evidence
       to support this should be included (do not send original
       paperwork).  All information requested below must be
       provided to enable an investigation."
           And then we've got the reference to debt recovery
       being on hold for seven days.
           If we look at what's required below: date of
       discrepancy, product, NBSC ref number, et cetera.  It is
       quite possible, isn't it, that somebody when they have
       come to settle centrally doesn't exactly know the date
       of the discrepancy or the product?
   A.  So if they don't know -- if they have settled an unknown
       branch discrepancy, we would immediately redirect them
       to NBSC.  We wouldn't -- we cannot pass the dispute on
       to a team to investigate if we don't have a team to pass
       it on to.
   Q.  So if somebody writes here "£2,000 short, must have
       happened over these three days, don't know exactly why",
       you can't deal with that?
   A.  No.
   Q.  And so is the debt unblocked?
   A.  We would call the branch back, or reply via email and
       tell them to ring NBSC for guidance on how -- to see if
       they can find the loss.
   Q.  But is the debt unblocked?  Is debt recovery --
   A.  It depends.  We're waiting for them to call NBSC and get
       a reference number, we then check the reference number
       at a later date and see what NBSC have advised.
   Q.  And if ultimately the person is not able to provide
       a specific date and product, will the debt be unblocked?
   A.  Yes, I'm afraid so.
   Q.  And why is an NBSC reference number required on the
       branch dispute form?
   A.  If they have one, we need it.
   Q.  It doesn't quite say that though, does it?  It suggests
       that it is a requirement that you have an NBSC reference
       number?
   A.  Well, if they've got an issue and they've got a loss,
       I would have expected them to ring NBSC prior.
   Q.  Even if they have settled it on the Friday, you sent
       your letter on the Monday ...?
   A.  Like I say, they still need to -- if they don't know
       what it's for they still need to ring NBSC.
   Q.  And what does your team do now you've got this
       collection of branch dispute forms?  You have told us it
       is about five a month.  Is there any analysis, or do you
       group them together, look for themes?
   A.  We don't deal with the actual dispute.  We pass them on
       to the relevant team.  That's why we need a product or
       something to pass on so we know which team to pass it on
       to deal with.  We're not experts on the transactions,
       but we will pass it on and get a response.
   Q.  And do you then track how they are resolved?
   A.  Yes.
   Q.  So do you know of the branch dispute forms that came in
       last year how many ultimately ended up in the
       subpostmaster repaying?
   A.  Not off the top of my --
   Q.  You don't track that?
   A.  No.
   Q.  Sorry, I spoke too quickly.
   A.  No, we don't track it.  Once the debt is cleared it is
       off the system and it's not part of the measures any
       more.
   Q.  But you appreciate it might ultimately never have been
       a debt?
   A.  Yes, and if something arises -- if they have paid and
       a transaction correction comes through after, they are
       welcome to keep the transaction correction that relates
       to the debt that they have already paid.
   Q.  So if there's a need to obtain transaction data or make
       an ALQ request, anything like that, that's not dealt
       with by your team?
   A.  No.
   Q.  Could we look please at {H/173/5}.  This is part of
       a letter sent by the claimants' solicitor to
       Post Office's solicitors on 22 January of this year.  Do
       you see at point 9 at the top it says the claimants'
       solicitors identified 11 to 12 in your witness statement
       and said:
           "Please disclose documents relating to Post Office's
       procedures in its use of the branch dispute form,
       including recording internal actions taken on receipt."
           Were you told about that request?
   A.  I think I was asked if there was any formal dispute --
       any process relating specifically to the dispute form,
       but it falls in with the rest of the dispute, it's just
       a process that we follow.  There's no formal policy or
       process in place.
   Q.  No formal policy or process in place ..?
   A.  Specifically to the branch dispute form.  It's generic
       to any dispute.
   Q.  So if we go just quickly to {H/227/3}, this is the
       response we see from 27 February 2019.  Do you see at
       point 9 again:
           "Post Office has confirmed that there is not
       a specific document which relates ..."
           Do you see that?
   A.  Yes.
   Q.  That's effectively what you have just described now?
   A.  Yes.
   Q.  And then it says:
           "A flowchart of the dispute process is at ..."
           And there is a specific number given.  Do you know
       what that flowchart is?
   A.  Not without looking at it, no.
   Q.  Can we go to {F/1276}.  Do you recognise that flowchart?
   A.  Not the title.
   Q.  There are two pages to this document, if you want to
       just glance quickly at page 2?
   A.  Not just yet, thank you.
   Q.  No.
           (Pause).
   A.  Okay, page 2 please {F/1276/2}.
   Q.  Just so you can see page 2.
   A.  That is a very old one.
   MR JUSTICE FRASER:  Very old did you say?
   A.  Yes.  It refers to a relationship manager which we do no
       longer ...
   MR JUSTICE FRASER:  Do you recall when you stopped having
       that?
   A.  2015 I think.  I can't be sure -- I'm not entirely sure
       when they left.
   Q.  Go back to page 1 please {F/1276/1}.  Do you see in the
       bottom half of the page there is a rectangle, "Issue
       passed to branch support team".  Does that still exist?
   A.  We've got a support services team.
   Q.  Not a branch support team?
   A.  That's not a name I would recognise them by: support
       services resolution team.
   Q.  So this is obviously not the current process, is it?
   A.  No, I just said it's not.
   Q.  If you go back over to page 2 please {F/1276/2}, who was
       the relationship manager?
   A.  Andy Winn.
   Q.  How long did he hold that position?
   A.  I don't know.
   Q.  And previously the process would be that it was
       Andy Winn who was essentially dealing with the
       resolution of disputes?
   A.  I can't answer to what the process was.  I know he dealt
       with disputes but I don't know at what point they went
       to him.
   Q.  Just on the face of this document it appears that he is
       the start of the dispute resolution process.  If you go
       back to page 1 --
   MR JUSTICE FRASER:  Ms Donnelly, I'm not sure that is really
       a question for the witness.  She said it is an old
       document, she said who the person was and she doesn't
       know.  Beyond that it is probably submission, isn't it,
       for me?
   MS DONNELLY:  Yes, thank you, my Lord.
   MR JUSTICE FRASER:  So back to you.
   MS DONNELLY:  Is there currently an appeal process?
   A.  In what respect?
   Q.  So you receive a branch dispute form.  You say it is
       sent off to the relevant team to look at and you say you
       track that.
   A.  Yes.
   Q.  If a subpostmaster is not satisfied with how that team
       resolved their dispute, is there an appeal?
   A.  They can come back to us and request and we will go to
       support services for assistance.
   Q.  That doesn't sound like an appeal process.  There's not
       another level?
   A.  For us, no, it's not.  Support services provide extra
       assistance for us.
   Q.  You are also the team leader for Santander banking.
       Just a few questions if I may in relation to some
       figures which appear in Mr Smith's witness statement.
           If we could go please to {E2/9/4}.  We see here this
       is part of the witness statement from Mr Smith that was
       made on 16 November of last year and do you see just at
       the top of the page it is actually the second part of
       paragraph 16 which starts on the previous page.  He
       says:
           "However, I have spoken to various team leaders
       within the business in order to gain a sense of
       proportion of TCs that have been challenged in their
       departments historically and what proportion of those
       challenges were successful (ie leading to a change in
       the TC before eventually being accepted or
       a compensating TC being issued).  The following sections
       of this witness statement are based on information
       provided to me by the team leaders identified in
       the headings ..."
           Then if we go to page 5 of the document {E2/9/5} you
       see at the bottom of the page "Agent accounting team",
       Dawn Phillips, that's you?
   A.  Yes.
   Q.  "Santander manual deposits" and we see he says:
           "In relation to Santander manual deposits
       Post Office received the following number of disputes in
       the 2016/17 financial year."
           And we can see the figures that appear there: errors
       received from Santander, 19,000-odd; disputes received
       from branch, et cetera and then it goes on.
   A.  Can I just say not all those disputes were from branch.
       They were not issued as transaction corrections.
       A dispute could have been sent on the basis of evidence
       received from the branch relating to a letter.
   Q.  If we just take it one step at a time for now.  Is it
       right that you provided those figures to Mr Smith?
   A.  Yes.
   Q.  And did you know the reason he wanted them?
   A.  Yes.
   Q.  So you knew that he wanted to give a sense of proportion
       of TCs that had been challenged?
   A.  Yes.
   Q.  And where did you get this information from?
   A.  From spreadsheets that we maintained on the team.
   Q.  So you had an existing spreadsheet with this information
       in?
   A.  Yes.
   Q.  Because how this document was read and understood by
       both experts acting in this case is that the first
       column was TCs issued by Post Office to branch?
   A.  No.
   Q.  And the second column is TCs disputed by branch to
       Post Office?
   A.  No.
   Q.  That's completely wrong?
   A.  Yes -- well, no, your perception is wrong I'm afraid.
       It does say errors received from Santander, it does not
       say TCs issued to branch.
   Q.  But you might not know that both experts read it in that
       way given the purpose that Mr Smith had set out earlier:
       that's what he said he was going to provide and this is
       what then appeared?
   A.  That is errors received from Santander, that is not
       transaction corrections issued to branch.
   Q.  And we see if we go to {E2/15} that Mr Smith provided
       a further witness statement.  We received this on the
       claimants' side at 17.48 on Friday 8 March and he says
       there that he has rereviewed his witness statement, do
       you see that within paragraph 3?
   A.  Yes.
   Q.  So he spoke to you again, did he?
   A.  Yes.
   Q.  When did you first find out that there was an issue with
       the figures you had provided that were in Mr Smith's
       witness statement?
   A.  It's not an issue with the figures, the figures remained
       the same, I think it is an explanation of what
       the headers actually mean.
   Q.  When did you first become aware that some further
       explanation might be required?
   A.  Because someone assumed that they were transaction
       corrections issued to branch and it was queried.
   Q.  Sorry, when did you become aware that this --
   A.  Not long ago.
   Q.  Because if you had provided this evidence rather than
       Mr Smith you might have been able to explain it a little
       bit more clearly, do you think?
   A.  Yes.
   Q.  And if we could go over the page please {E2/15/2}, at
       paragraph 5 he says:
           "The figures in my first statement of 2,890
       'disputes received from branch' represents the number of
       disputes between Post Office and Santander ..."
   A.  Yes.
   Q.  And then paragraph 6:
           "I understand from my colleague Dawn Phillips that
       the Santander manual deposits team issued 3,968
       transaction corrections during the financial year
       2016-2017."
           So if someone had misread Mr Smith's first witness
       statement, we've got about a fivefold difference from
       about 19,500, it's actually only around 4,000?
   A.  That were issued.
   Q.  Now, what he hasn't told us, or isn't clear from his
       witness statement, perhaps you can just clarify, is how
       many of the 3,968 transaction corrections that were
       issued were disputed?
   A.  I can't answer that.  I have already been asked that
       question and it's not something we recorded as to
       whether Post Office disputed it with Santander on the
       postmaster's behalf with evidence from the branch, or it
       was an actual transaction correction dispute.
   Q.  Had Mr Smith specifically asked you that?
   A.  Yes.
   Q.  And you told him it's not something that --
   A.  It's not something we recorded.
   Q.  Could we just go back please to {E2/9/5}.  Could you
       just help us understand if about 19,500 errors are
       received by Post Office from Santander and Post Office
       actually issue only around 4,000, what has happened to
       the rest?
   A.  So Santander do a manual match of figures to Horizon to
       the actual paperwork that they receive.  If it doesn't
       match, they sent us an error notice.  They were all in
       different categories, so several of the categories we
       know it was a timing issue and if we just waited a week
       they would send us another compensating error to clear
       it.  A lot, I would say 50% of what they sent us could
       be cleared with a letter to branch requesting copies of
       their paperwork which we could then send back to
       Santander.
   Q.  So it's quite a lot more complicated than the way it is
       presented perhaps in Mr Smith's first or second
       statement?
   A.  Well, maybe.  We issued a lot less than Santander gave
       us.
   Q.  So you are doing some sort of internal --
   A.  Yes.
   Q.  -- reconciliation yourself?
   A.  Prior to --
   Q.  That's a manual process, is it?
   A.  Yes.
   MS DONNELLY:  I don't have any further questions, my Lord.
   MR JUSTICE FRASER:  Thank you very much.
           Mr De Garr Robinson, any re-examination?
              Re-examination by MR DE GARR ROBINSON
   MR DE GARR ROBINSON:  I have only one question, Ms Phillips.
       You were asked about the dispute process and you were
       asked what happened if the subpostmaster was not
       satisfied with a decision that was made at the end of
       that process, was there an appeal, and you indicated
       that you would then call in support services for
       assistance.  Could you explain what support services
       would do?
   A.  They are capable of doing a more in-depth investigation
       than perhaps we could in FSC, so they would be able to
       provide more information and a breakdown of what
       happened in branch.
   Q.  And how detailed would their investigation be?
   A.  Extremely detailed.
   Q.  Thank you, Ms Phillips.
           My Lord, I have no further questions.
                 Questions from MR JUSTICE FRASER
   MR JUSTICE FRASER:  Just a couple of questions on the
       table -- could we go back to it please {E2/9/5}.  We've
       got E2/9/5 on the screen which you have been asked some
       questions about.  The first column I think you say the
       number is correct but the description is wrong; is that
       fair?
   A.  No, the errors are received from Santander, but they are
       not transaction corrections issued to branch.
   MR JUSTICE FRASER:  Should I reading the heading to column 2
       differently?
   A.  Yes, as in Paul's second statement --
   MR JUSTICE FRASER:  Well, I'm rather asking you because you
       provided the figures.
   A.  So they are just disputes sent to Santander from Paul.
       They can be disputes where we have sent a letter to
       branch requesting evidence and not issued a transaction
       correction, we have just asked for the evidence and it
       has been provided, or they can be transaction correction
       disputes.
   MR JUSTICE FRASER:  And as far as you know are the headings
       to the other three headings correct?
   A.  Yes.
   MR JUSTICE FRASER:  They are.
           One question Ms Donnelly asked you in respect of
       this and the total which I think is from Mr Smith's
       second statement is 3,968 in terms of transaction
       corrections.  She asked how many of those 3,968
       transaction corrections that were issued were disputed
       and you said you can't answer that, you have already
       been asked that question and it's not something you
       record.
   A.  No.
   MR JUSTICE FRASER:  When were you asked that question?
   A.  In recent weeks.
   MR JUSTICE FRASER:  In recent weeks leading up to this
       trial?
   A.  Yes.
   MR JUSTICE FRASER:  You weren't asked it at the time in
       respect of these figures?
   A.  Not the first statement.  That's how the second
       statement came around because we were asked to -- we
       were asked that question: how many of those TCs were
       actually disputed by branch, and I can't answer it.
   MR JUSTICE FRASER:  Understood.
           I assume nothing arising out of that?
           Thank you very much for coming.  You have a much
       shorter stint in the witness box than
       Mrs van den Bogerd.
   MR DE GARR ROBINSON:  My Lord, I wonder whether this would
       be a convenient moment for a break.
   MR JUSTICE FRASER:  I think it would.  Who is your next
       witness?
   MR DE GARR ROBINSON:  It is Mrs Mather.  I hope I'm
       pronouncing her name correctly.  It is one of the things
       I want to check.
   MR JUSTICE FRASER:  Let's ask her right at the beginning.
       And is there a fighting chance, Mr Green, of finishing
       her evidence today or is she going to go into tomorrow?
   MR GREEN:  My Lord, we're hoping to finish her and Mr Smith
       if we can.
   MR JUSTICE FRASER:  Excellent.
   MR GREEN:  Definitely her.
   MR JUSTICE FRASER:  None of this is binding it is just to
       give me a general outline.
   MR GREEN:  Of course.
   MR JUSTICE FRASER:  Okay, we will come back in at 20 past.
   (3.10 pm)
                          (Short Break)
   (3.20 pm)
   MR DE GARR ROBINSON:  My Lord, I call Tracy Mather -- that's
       how one actually pronounces her name.
                     MRS TRACY MATHER (sworn)
   MR JUSTICE FRASER:  Thank you, Mrs Mather.  Do have a seat.
           Examination-in-chief by MR DE GARR ROBINSON
   MR DE GARR ROBINSON:  Now, Mrs Mather, in front of you there
       should be a file of documents.  I would like you to go
       to tab 8 of the file please {E2/8}.  Do you see --
       I think there's a sheet of corrections, but after that
       sheet of corrections do you see a witness statement of
       "Tracy Jane Wendy Mather"?
   A.  Yes.
   Q.  Is that your name and address on the first page?
   A.  It is.
   Q.  If you go to the back of the document, is that your
       signature?
   A.  It is.
   Q.  And do you also see a document with three corrections to
       your witness statement at the front of that file?
   A.  I do.
   Q.  And now, Mrs Mather, I have to ask a question which is
       born out of sheepishness on my part.  Everyone else in
       court will understand why, but you won't.  I would like
       to ask you about paragraph 12 of your statement please
       {E2/8/3}.  Here you are describing Credence and you say:
           "Credence is used as an information tool.  It is
       designed to work alongside other applications.  It is
       used to help understand what has happened in a branch as
       it records all keystroke activity performed in that
       branch by the user ID, date and time."
           Could I ask you to explain what you mean by the
       phrase "It records all keystroke activity"?
   A.  What I actually meant was the transactional data as in
       sales and non-sales.
   Q.  I see, thank you.
           Subject to the three corrections you have identified
       and your clarification of paragraph 12, do you confirm
       that this witness statement is true to the best of your
       knowledge, recollection and belief?
   A.  It is.
   Q.  Thank you, if you could wait there please.
                  Cross-examination by MR GREEN
   MR GREEN:  Now, Mrs Mather, at paragraph 1 of your witness
       statement, if we look please {E2/8/1}, you say you are
       the FSC team leader, so are you the team leader for FSC?
   A.  No, sorry, for my team.  It's just a generic team leader
       role in FSC.
   Q.  You are a team leader within FSC?
   A.  Yes.
   Q.  And which is your team?  Have you got a special team
       name?
   A.  For the last year and a half I have had MoneyGram and
       cheques I have had for two years, just over two years.
       Before that I was on payout, postal orders.
   Q.  Different things?
   A.  Yes.
   Q.  And how many people work in FSC overall?
   A.  I would be guessing.
   Q.  Roughly?
   A.  60?
   Q.  And do you have any special training when you move to
       a new product you are dealing with?
   A.  Not training of such.  You tend to -- because the team
       leaders might move on to another team, so you would
       spend time with that team leader and then when they move
       into a new area you can always go back if there was
       something you were unsure of.
   Q.  And when you join FSC in the first place do you sort of
       pick it up on the job, how does it work?
   A.  FSC -- I have always been part of that kind of group
       since I joined.
   Q.  Did they give you any special is training when you
       joined, or did you ..?
   A.  We're going back a long time, but my special training of
       such was when I was a PA so I was machining so yes, they
       would give me special training.
   Q.  The purpose of your statement I think, as you explain at
       paragraph 8 {E2/8/2}, is to explain how Post Office uses
       Credence to investigate discrepancies.
   A.  Yes.
   Q.  "... since the introduction of Horizon and why".  And at
       paragraph 9 you have explained that:
           "Credence is a system that shows all sales
       transactions, log on and log off events and all
       remittances in and out of Horizon."
           Yes?
   A.  Yes.
   Q.  And we have seen already that paragraph 12 of your
       witness statement {E2/8/3} says it is used to help -- if
       you look at the second line, that Mr De Garr Robinson
       just took you to:
           "It is used to help understand what has happened in
       a branch as it records all keystroke activity performed
       in that branch by the user ID, date and time."
           Now, you have said that what you meant by that in
       fact was sales and things like that?
   A.  Transactional data, yes.
   Q.  What information does it give above and beyond the ARQ
       data that we have seen?
   A.  We -- for my team, which is the cheques team, where we
       usually use Credence, is years ago we would have had to
       ask a postmaster to send his cheque list in for us, if
       we've got an open item to investigate.  With Credence we
       can save not going out to the postmaster, we can
       actually have a look to see how postmasters finish
       transactions off.
   Q.  And what he has typed in in the cheque list at the
       branch and how it shows up or --
   A.  No, we actually look at all sales to see what
       transactions he has finished off to cheques.
   Q.  So you don't have access to what we have seen is the ARQ
       data that's being pulled from Fujitsu?
   A.  Mm-hm.
   MR JUSTICE FRASER:  Is that correct, you don't have access
       to that?
   A.  Not through Credence, no.  We're just at the
       transactional level.
   MR GREEN:  So can you explain to the court -- give the court
       some understanding of what you can and cannot see.  If
       you're investigating, for example, a MoneyGram issue,
       can you see exactly what the SPM has done in branch on
       the screen in Credence?
   A.  With MoneyGram we have two sources of information and
       that is MoneyGram's file and we use Horizon, we don't
       actually use Credence for MoneyGram.
   Q.  Okay.  What do you use Credence for?
   A.  Mainly cheques, part of my team.
   Q.  Now, Credence only holds information for a short period,
       I think currently three months; that's right, isn't it?
   A.  That's right.
   Q.  And are you aware of discussions that have been had
       about extending the period for which Credence holds
       information to a longer period?  Are you aware of any
       discussions about extending the period that Credence
       holds information for?
   A.  We know that we can view Credence up to three months.
       There's also a system called Horace.  So we can -- one
       of my team members has access to Horace, so we would
       look on there.
   Q.  Okay.  Let's look if we may at {F/1092} please.  Now,
       this is a document from June 2013.  You will see as at
       2013 the branch data available as at that date was
       available for 61 days.  Is that the same today as far as
       you know?  Branch data available to you on Horizon?  Do
       you see at the top?
   A.  Yes, I thought ours was 90 days, Credence.
   Q.  Well, Credence is immediately underneath, Credence data
       90 days.
   A.  Yes.  But when you say branch data I'm not too sure --
   Q.  On Horizon.
   A.  Yes.
   Q.  How long is the data available to you on Horizon for?
   A.  On Horizon, if we have an open item it's there until we
       investigate it.
   Q.  It can just stay on there for a long time?
   A.  Yes.
   Q.  And next to that it says:
           "All transactional data including non-sales eg
       adjustments."
           So it's right, isn't it, that Credence doesn't just
       show sales information, it also includes non-sales
       information like any adjustments that somebody has made?
   A.  Yes.
   Q.  And so that gives a fuller picture --
   A.  Yes.
   Q.  -- than just seeing the sales information?
   A.  Yes.
   Q.  And that's why in your statement you have explained that
       it helps to see what's happened in the branch?
   A.  That's it.
   Q.  And can you see that the -- under -- you can see what's
       being considered there in bold.  Do you see about
       quarter of a way down from the top in bold:
           "Consider the benefits of a repository of data for
       24 months.  Note: this is open for discussion but
       24 months is the initial gambit."
           Do you see that?
   A.  Yes.
   Q.  Would it be more helpful to you to have Credence data
       available for longer?
   A.  We have -- there are some reports that we can still see
       information for 24 months on Credence.
   Q.  Are there now?
   A.  Yes.
   Q.  Do you know when that was extended up to 24 months?
   A.  I don't, no, because I wasn't on these particular teams
       at that time, because in 2003 I weren't on my job.
   Q.  You don't have to know the answers to the questions,
       it's fine, but if you do it is helpful for the court to
       know what the answer is.
           Under "Business benefits" it says:
           "The business, in particular the security and P&BA
       departments suffer from the inability to interrogate its
       data due to the short periods of retention."
           Now, P&BA, what is P&BA now called?
   A.  It is now FSC.
   Q.  So that's your department that you work in?
   A.  Yes.
   Q.  And we can see that what's suggested there is "the
       following areas should benefit", "POCA claims and
       disputes".  Can you tell the court what POCA is?
   A.  I don't know.
   Q.  If we go down and look at "Fraud and conformance
       investigations", do you see that?
   A.  Yes.
   Q.  Is that something you have ever been involved in or not?
   A.  No.
   Q.  And then "SPMR contract advisors - re non-conformance
       suspension hearings".  Have you ever been involved in
       anything like that?
   A.  No.
   Q.  And then if we go down, just underneath "Financial
       benefits" you will see:
           "ARQ (audit retrieval process) costs at least 384k
       recurring annually.  This is subsumed without breakdown
       in the Fujitsu security management costs.  720 requests
       @ £450 unit cost."
           I think you said earlier you don't generally get the
       ARQ data, is that right?
   A.  No.
   Q.  Have you ever worked in a team where someone has sought
       the ARQ data?
   A.  I would imagine it's when it's the fraud team or the
       security team.
   Q.  Something like that?
   A.  And they need more data.
   Q.  So we know that at least some reports on Credence are
       now available for 24 months?
   A.  (Nods).
   Q.  Can you just give the court an indication of some
       examples of those?
   A.  It's usually regarding the automated payment, bill
       payments.
   Q.  Automated bill payments?
   A.  Yes.
   Q.  So those reports are available for longer?
   A.  Yes.
   Q.  Are there any others that you can think of?
   A.  Not off the top of my head.
   Q.  Don't worry.  It's not Mastermind.
           Now, can I ask you please to be shown {F/1120}.
       This is a PEAK document about the Horizon computer
       system.  Do you ever see these in the course of your
       work or not?
   A.  Me, no.
   Q.  Does anyone in your team ever have cause to look at
       a PEAK?
   A.  No.
   Q.  No?  Can I just ask you then -- I will take it quite
       quickly, but if we look at the bottom of that first page
       you will see the bottom paragraph underneath the word
       "Details":
           "The customer visited Kinglassie post office ... on
       13 December 2012 ... to deposit 1,444.20 which didn't
       credit their account.  We can see from TESQA."
           What's TESQA?
   A.  I've never worked on this team, but I'm aware that it's
       another system that people -- there's more details on
       that.
   Q.  As well?
   A.  Which is more restricted obviously for data protection
       so it is only certain people that have access to that
       data.
   Q.  And that's a separate system to Credence?
   A.  Yes.
   Q.  "We can see from TESQA that the transaction failed at
       the confirmation stage due to 'transaction abandoned by
       clerk'.  I've had numerous conversations with the office
       and also sent out a letter and evidence but the office
       has now requested a transaction log.  With it being so
       long ago they can't provide an explanation as to why the
       transaction didn't complete (but state they didn't have
       a surplus at the time of the transaction).  If we could
       ascertain what buttons were pressed by the clerk we
       could go back with an explanation as to what the clerk
       did with this transaction and hopefully get a conclusion
       to this enquiry."
           Do you see that?
   A.  Yes.
   Q.  Now, it would be helpful, wouldn't it, when you have
       those sorts of issues if you could identify as much
       information about exactly what had happened in branch as
       possible?
   A.  I can't answer for that particular enquiry.  On cheques
       we have the information that we require to investigate.
   Q.  You do?
   A.  Yes.
   Q.  Even if it is a long time afterwards?  If someone raises
       it say 102 days later for the first time?
   A.  If we issued a transaction correction to the office, we
       put our details on there and if they had an enquiry or
       dispute something they would phone the duty (inaudible)
       up for more information.
   Q.  But I think what has happened here, on the face of it --
       and this is not a cheque case, but what appears to have
       happened here is that it has been raised -- can you see
       that the incident is raised on 30 August 2013?
   A.  Yes.
   Q.  And it relates to a visit on 13 December 2012, do you
       see that?
   A.  Yes.
   Q.  So we can see that more than 90 days has elapsed from
       the visit of the customer to the time when the issue has
       arisen and if you looked on Credence you wouldn't be
       able to see the details of the original visit, would
       you, after longer than 90 days?
   A.  No.
   Q.  Do you know how long you can see the details on TESQA
       for?
   A.  No.
   Q.  You don't.  Okay.
           Now, if you just go over the page at {F/1120/2}, you
       see at the top of the page:
           "We need sight of the actual transaction retrieving
       from the archives (this is required to disprove any
       future claim against Horizon integrity via the clerk)."
           Was it your experience at all that there was any
       defensiveness about the Horizon system when it was
       challenged by SPMs?
   A.  If a postmaster challenged in the area that I worked in,
       we always took them seriously --
   Q.  You did?
   A.  -- and we always looked at what they were actually
       saying and then we would go back to our evidence of
       issuing the TC.
   Q.  Okay.  Can we look at your paragraph 13 please {E2/8/3}.
       You can see there, third line:
           "For example, a subpostmaster might telephone FSC
       and/or the helpline and allege to Post Office that
       he/she has done a reversal."
           Yes?
   A.  Yes.
   Q.  That's an SPM claiming to have done something.  The way
       that is expressed suggests some doubt a little bit,
       doesn't it, about the truth of what they're saying?
   A.  I think what I was trying to get across in this
       particular statement was if a postmaster had a query, he
       could always phone us and we would try and help,
       whatever his query was.  That's what I was trying to
       explain in this.
   Q.  Okay.  Let's look if we may at paragraph 14.  You say
       there:
           "In relation to Mr Coyne querying whether Credence
       could contain errors that result in false TCs being
       issued to branches which they cannot effectively
       challenge, as I am an end user of Credence I cannot
       really comment on this.  However I've never heard of
       a bug in Credence in my time at Post Office."
           So that's not something you are in a position to
       comment on?
   A.  No.
   Q.  Okay.  Let's move forward if we may please.  Let's look
       please at {F/869}.  This is an Ernst & Young management
       letter for the year ended 27 March 2011 and I expect you
       may not have seen that before, is that fair?
   A.  That's fair.
   Q.  Can we go to page 4 please {F/869/4} and just to show
       you the layout of the table, the first column has
       a number in it, the second column says "Issue", although
       it is a bit strangely laid out, do you see?
   A.  Yes.
   Q.  "Location", "Background", "Recommendation", "Management
       comment" and then the right-hand column is "Current year
       update", do you see that?
   A.  Yes.
   Q.  And let's look at page 12 at point 12 please {F/869/12}.
       Point 12 says "Credence (back-end) change process" and
       you can see there it says:
           "During our walkthrough and testing of the change
       control procedures for the Credence application we
       became aware of the following issues:
           "1.  Developers at Logica, the third party provider
       of application development and support for Credence, had
       access rights to the production environment and the
       database that would permit developers to move their own
       changes into the production environment.
           "2.  Documentation to approve fixes and patches that
       are applied to Credence outside of the release process
       does not always exist.  We were advised by Logica
       personnel that for a sample of four changes selected
       evidence of approval to move into production did not
       exist and that it would not be possible to link the
       changes to problem tickets to record the original
       request for the fix ..."
           Then:
           "Developers have access to move their own changes
       into production and [if] documentation is not retained
       to substantiate those changes there is a risk of loss of
       data and application integrity due to either
       unauthorised, erroneous or inappropriate changing made
       to the production environment."
           I'm not suggesting somebody read anything as
       complicated as that out to you, but were you aware at
       all in or around 2011 or thereafter of any concerns
       about the integrity of the information on Credence?
   A.  No.
   Q.  Okay.  Let's look at page 14, point 13 {F/869/14}.  This
       is "Credence (front end) change process" and we can see
       some issues identified in the column where there is
       a number 137 and in fact over on the right-hand side it
       says:
           "Application not in audit scope for [financial
       year 11].  Therefore we are not able to comment on
       whether management has fully addressed our comment as
       raised in the prior year."
           You didn't see any particular changes in Credence
       around 2011 or soon thereafter?
   A.  To be honest, I wasn't working in this area in 2011.
   Q.  What area were you -- were you using Credence at all at
       that time?
   A.  Not at that particular time, no.
   Q.  You weren't, okay.
           Before reading Mr Coyne's report were you aware that
       there is a time discrepancy between the Credence and the
       Fujitsu data?
   A.  What do you mean?
   Q.  One is GMT and one is British Summer Time for half the
       year.  Did you know about that?
   A.  No.
   Q.  No.  Fair enough.
           Let's look at the use and content of Credence data.
       If we look at {F/935}, and you can see that that
       document is called "Driving business benefits through
       the consolidation of data review" and dated 18 May 2012.
           Can I just ask you at what date did you move into
       a role where you started using Credence yourself?
   A.  Probably 2016.
   Q.  Okay, so this is --
   A.  '17, yes.
   Q.  -- before your time as well, okay.
           So I will leave that.
           Paragraph 15 of your witness statement {E2/8/3}
       says:
           "Looking at the Helen Rose report referred to in
       paragraph 5.49 of Mr Coyne's report, Post Office was
       able to use Credence to identify that the subpostmaster
       had reversed a transaction but had also taken £76.09
       payment from the customer."
           And you have I think made an amendment to that and
       if we look at E2/16 and we go forward one page {E2/16/2}
       you can see that paragraph 15 now reads -- "Post Office
       was able to use Credence" should now be "Post Office
       initially used Credence".  And then there is the
       addition of a sentence at the end that says:
           "However, page 2 of the Helen Rose report shows that
       the reversal 'was not an explicit reversal by the
       clerk'."
           Now, first of all, you are talking about the
       Helen Rose report; is that something that you actually
       knew about yourself?
   A.  No.  In this particular incident though when there was
       not an error in the core finance, why this came about
       was because the customer contacted the helpline because
       the bill was unpaid.
   Q.  Yes.  And if we look at how your paragraph 15 naturally
       reads {E2/8/3}, if we look at the last sentence there,
       do you see it says:
           "In reversing the transaction, the subpostmaster had
       effectively removed the payment to British Telecom,
       making the bill unpaid."
   A.  Yes and it was the system that removed it, not the
       postmaster, yes.
   Q.  It wasn't the subpostmaster.  You hadn't appreciated
       that from the Helen Rose report --
   A.  No.
   Q.  -- because the Helen Rose report wasn't really something
       you knew about?
   A.  No.  But it was the recovery process that reversed --
   Q.  Could have done it?
   A.  Yes.
   Q.  So that amendment has now been made to reflect that?
   A.  Yes.  And the transaction correction was issued because
       the customer's money was still in the branch, so by
       issuing the transaction correction the AP team was able
       to then repay the customer's bill.
   Q.  And at paragraph 18 {E2/8/4}, somehow ARQ requests has
       come into your witness statement.
   A.  Yes.
   Q.  It's not really your thing, is it?
   A.  No.
   Q.  Not really.  I want to be fair to you about how I ask
       it.  It says there:
           "I understand that Mr Coyne has alleged that
       Post Office staff were deterred from making ARQ requests
       because of fees or penalties."
           Yes?
   A.  Yes.
   Q.  Did someone point that out to you in Mr Coyne's report,
       or did they just tell you that that's what he had said?
   A.  I have been always aware that Post Office Limited teams
       can ask for this information, because many years ago we
       didn't have Credence and we would ask for these, but
       I've never been aware of anybody being deterred from
       making requests.
   Q.  But no one in your team actually does that?
   A.  No, because we have the information at hand that we
       require.
   Q.  And you can see things on Credence that the SPM can't
       see and so you try and deal with their queries?
   A.  We can see the same information.  Ours is probably more
       easily readable.
   Q.  In a spreadsheet form?
   A.  In a spreadsheet format.
   Q.  Not on a till roll?
   A.  Not on a till roll.
   Q.  And you say that you have spoken to Mr Godeseth.  He is
       actually giving evidence, isn't he?
   A.  As I have put "I understand from a witness statement",
       I didn't actually speak to him.
   Q.  I'm so sorry, you are right, you say "I understand the
       witness statement contains details".
   A.  Yes.
   Q.  Then you say -- the reason I thought you might have
       spoken to him is three lines from the bottom, "He is not
       aware of any fees or penalties having been paid".  Is
       that just from his witness statement --
   A.  That's it, yes.
   Q.  -- you haven't spoken to him?
           Can we look please at {F/728}.  This is an email
       chain in relation to the Barkham Post Office, which was
       Pam Stubbs' Post Office, and if we go please to page 11
       {F/728/11}, you can see there:
           "If we can encourage Mark Dinsdale to authorise the
       audit trail, I feel it would be beneficial given the
       current interest in Horizon from media and MPs."
           If we go to page 9 {F/728/9}, we get:
           "Nigel, no probs with requesting data from Fujitsu
       but it will take around 3 weeks.  Has Jason agreed to
       take this case on, because we don't hand over Horizon
       logs to an SPMR.  It needs an expert to understand what
       it says, and usually this requires one of the
       investigators.
           "I'll give Jason a call in the morning, then I will
       raise an ARQ from Fujitsu.
           "Is this for our benefit, as there is a cost
       attached to ARQ requests, we do get a supply free of
       charge as part of the contract but we usually don't have
       enough, therefore we usually charge the defence
       lawyers."
           Now, does that chime with anything that you have
       experienced?
   A.  No.
   Q.  Let's have a look, if we may -- well, I may not need to
       take you to it, but you're not aware of the charging
       regime at all yourself, are you?
   A.  I was aware Post Office had a certain amount and then
       they could be chargeable.
   Q.  So you were aware they had a certain amount after which
       they would be charged?
   A.  Yes.
   Q.  But you didn't know yourself what that amount was?
   A.  The charge?  I have no idea, no.
   Q.  And you didn't know how many requests they could make
       before they were charged?
   A.  No.  If we needed to make a request, I knew we could.
   Q.  I'm grateful.
           My Lord, I have no further questions.
   MR JUSTICE FRASER:  Mr De Garr Robinson?
              Re-examination by MR DE GARR ROBINSON
   MR DE GARR ROBINSON:  Mrs Mather, I just have a question
       about paragraph 19 then.  It is a potential for
       confusion that you might be able to clear up.
       Paragraph 19 of your witness statement {E2/8/4}.
       Mr Green asked you about information you had obtained
       from Mr Godeseth and you said you had just read his
       statement.
   A.  Yes.
   Q.  And then you referred to some information at the end of
       that paragraph where it says:
           "He is not aware of any fees or penalties having
       been paid by Post Office, nor anyone being deterred from
       making ARQ requests ..."
           In that sentence who is "he"?
   A.  I think it's Christopher Knight.
   Q.  So it's not Mr Godeseth?
   A.  No.
   Q.  And is that information as a result of your having
       spoken to Mr Knight?
   A.  Yes.
   Q.  Thank you.
           My Lord, I have no further questions.
                 Questions from MR JUSTICE FRASER
   MR JUSTICE FRASER:  I just have a couple of questions.
           The one that you have just been asked by counsel
       about Mr Christopher Knight, is there any reason of
       which you are aware why you are telling us what
       Mr Christopher Knight believes rather than Mr Knight
       himself giving a witness statement?
   A.  I have no idea.
   MR JUSTICE FRASER:  You don't know.
           Then if we could go back please to your paragraph --
       well, it's really your experience of Credence
       altogether.
           How often do you use Credence?
   A.  My team use it every day.
   MR JUSTICE FRASER:  And how often do you use it?
   A.  Perhaps once a week.  I know my way round Credence.
   MR JUSTICE FRASER:  You know your way round it.
   A.  Yes.
   MR JUSTICE FRASER:  And what is keystroke activity?
   A.  I didn't realise that keystroke activity actually refers
       to these ARQ reports where we do the transactional data
       on Credence; that's what we're looking at, the sales and
       the non-sales.  So I meant transactional data.
   MR JUSTICE FRASER:  So your understanding of keystroke
       activity when you did your statement was what?
   A.  Transactional data, sales and non-sales.  So for every
       transaction that a postmaster -- for example somebody
       paid their bill, we would see that level of detail and
       how the transaction was finished, either to cash,
       cheque, debit card.  So it's the transactional data over
       the Horizon counter.
   MR JUSTICE FRASER:  Right, so if I look at your paragraph 12
       {E2/8/3} -- do you have that?
   A.  Yes.
   MR JUSTICE FRASER:  You see the sentence that you have been
       asked about a couple of times --
   A.  Yes.
   MR JUSTICE FRASER:  -- firstly by Mr De Garr Robinson and
       then by Mr Green.  Just so that I can understand what
       your evidence is, that second sentence that begins "It
       is used to help understand ...", I understand I should
       cross out "all keystroke activity", is that right?
   A.  Yes.
   MR JUSTICE FRASER:  And what term should I put in there
       instead?
   A.  What we use Credence for is we have also got -- for
       cheques is all cheques are processed at a processing
       centre which they then create a file.  So for example if
       we had ten cheques processed we would look on Credence
       for ten transactions that were finished off to cheques,
       so ten versus ten and then if we've got a mismatch
       that's what we would be looking for.  For example one
       could be finished off to cash accidently and we would
       try and identify that.
   MR JUSTICE FRASER:  But going back to my question, what term
       should I put in instead of "all keystroke activity"?  Is
       there a handy or useful form of words that would correct
       that sentence to reflect what you meant?
   A.  Records all transactional data performed in that branch.
   MR JUSTICE FRASER:  All transactional data performed in that
       branch.  Okay.
           And then the next question, which is more or less on
       the same topic, in the next paragraph you have
       a sentence that begins at the end of line 4 which says:
           "We can take an initial look on Credence to see
       whether that is correct or whether, for example, the
       subpostmaster has in fact done an adjustment."
           Do you see that?
   A.  Yes.
   MR JUSTICE FRASER:  Can you just explain to me how you would
       do that?
   A.  For example, if one of -- a postmaster phoned one of my
       team, they might think that they have done a reversal,
       but by the duty (inaudible) they would be able to go on
       to Credence, download the information and look on
       non-sales and they might have actually done an
       adjustment.
   MR JUSTICE FRASER:  And is that something that you would be
       involved in looking at, or would it be someone in your
       team?
   A.  It would be my team members.
   MR JUSTICE FRASER:  And could you do that if you wanted?
   A.  If I wanted to, I could, yes.
   MR JUSTICE FRASER:  And what would it show?  What would you
       expect to see in your example?
   A.  That's where -- for an adjustment it's when they can
       adjust cash to cheque or cheque to cash.
   MR JUSTICE FRASER:  Right.  That's very useful.  Thank you
       very much.
           Any questions arising out of that?  No.
           Thank you very much for coming, Mrs Mather, that's
       the end of your evidence.
   MR DE GARR ROBINSON:  My Lord, I call Paul Smith.
   MR JUSTICE FRASER:  Yes.
                     MR PAUL SMITH (affirmed)
   MR JUSTICE FRASER:  Thank you very much, Mr Smith.  Do have
       a seat please.
   A.  Thank you, my Lord.
           Examination-in-chief by MR DE GARR ROBINSON
   MR DE GARR ROBINSON:  Mr Smith, there should be a bundle of
       documents in front of you.  If I could ask you to open
       that bundle and go to divider 9 please {E2/9}.  There
       should be one sheet with a correction in it and after
       that sheet there should be a document describing itself
       as a "Witness statement of Paul Ian Michael Smith", do
       you see that?
   A.  Yes.
   Q.  And is that your name and address on the first page of
       the witness statement?
   A.  It is.
   Q.  And at the end of the witness statement -- it is page 7
       {E2/9/7} -- is that your signature?
   A.  It is.
   Q.  And then if I could ask you to move on in the bundle to
       what I hope is tab 15 -- is there a tab 15 in that
       bundle?  {E2/15}
   A.  There is.
   Q.  And another witness statement with your name and address
       on the first page and on the second page is that your
       signature?  {E2/15/2}.
   A.  It is.
   Q.  Subject to the correction on the first sheet relating to
       the first witness statement, do you confirm that the
       witness statements are true to the best of your
       knowledge, recollection and belief?
   A.  They are.
   Q.  If you would wait there, thank you.
                  Cross-examination by MR GREEN
   MR GREEN:  Mr Smith, we have heard how you discussed the
       information about transaction corrections with various
       colleagues and one of them was Dawn Phillips.
   A.  It was.
   Q.  And can I just take you please to your witness statement
       at {E2/9/3} at paragraph 13.  So you have introduced TCs
       at paragraph 12.
   A.  Mm-hm.
   Q.  And it says:
           "TCs are issued by FSC.  I understand from
       Post Office's solicitors that the processes by which FSC
       determines whether a TC is required is outside the scope
       of the Horizon issues trial."
   A.  Okay.
   Q.  That's what you were told?
   A.  Yes.
   Q.  And then you said:
           "Broadly, it involves FSC comparing data entered
       onto Horizon by branch staff with data received from
       other sources and resolving any discrepancies between
       them."
           Now, pausing there, there are -- I think as we have
       already heard -- separate teams for different potential
       discrepancies?
   A.  Agreed.
   Q.  And you have explained at paragraph 14 that:
           "FSC is divided into distinct teams that are
       responsible for accounts relating to specific products."
   A.  They are.
   Q.  So one team might be dealing with Santander
       discrepancies, for example?
   A.  Indeed.
   Q.  Another team might be dealing with lottery and so forth?
   A.  Absolutely.
   Q.  And you say that:
           "Although each team have their own responsibilities,
       there is interaction between teams to reach a final
       resolution on any discrepancy."
   A.  There is.
   Q.  When would that typically occur?
   A.  That would occur when there is evidence within the
       investigation that there could be a further activity
       that's needed.  As an example, we have already heard
       about cheques.  It could be that the impact of a cheque
       discrepancy could also impact on another product, so if
       you have a miskey it could be that that miskey
       transaction has been settled to cheque with higher than
       what it should be value, but then there's also going to
       be the bill payment or the actual transaction itself
       that's also been at a higher value.  So working together
       you can put together all the pieces of the puzzle, get
       the right answer and also mitigate or minimise the
       impact on the branch by making sure that any corrections
       that are done are done at the same time to ensure that
       the financial impact is either mitigated or reduced.
   Q.  Yes.  It is obviously ideal if both sides can be done at
       the same time.
   A.  Absolutely.
   Q.  And the subpostmaster is not hit with a TC going one way
       and then another way.
   A.  Why he.
   Q.  But that does happen sometimes?
   A.  I cannot give you an example --
   Q.  Okay.
   A.  -- but I can't deny it either.
   Q.  At paragraph 16 you say:
           "Post Office introduced a case management system
       that records each individual challenge to a TC ..."
           Yes?
   A.  It records some challenges to TCs.  It is still
       something that's in roll-out.  It was introduced on that
       date.  There are a number of teams that are on there and
       challenges to TCs are recorded in there, yes.
   Q.  So taking it in stages, it was introduced
       in September 2018?
   A.  Agreed.
   Q.  And when is roll-out of that system due to be complete
       across all teams?
   A.  I can't give you a date.  I don't know a date.
   Q.  Okay.  And how many teams are there in total?  Roughly?
   A.  Roughly I would say eight.
   Q.  Eight.  And do you know how many teams are using that
       system at the moment?
   A.  I know that there are two teams which are high volume
       users that are fully integrated in there and there are
       a number of teams which have maybe part of their
       processes in there.  As I explained previously, teams
       have distinct areas of responsibility.  It could be that
       a product is integrated but the whole team isn't.
   Q.  Okay.  Which are the two teams that are high volume
       users?
   A.  The automated payments team and the team that deals with
       debit cards, lottery and ATM.
   Q.  And do you know whether they are using it to record all
       challenges or just some challenges?
   A.  As far as I'm aware, all challenges.
   Q.  Okay.  And do you know whether that case management
       system actually tracks any trends or if there's any
       analysis of the figures?
   A.  The case management system itself doesn't because it's
       a case management system, it's there to record.
       However, there are reports being developed that are
       showing some trends.  I do know that the lottery, ATM
       and debit card team leader has requested a number of
       reports and is using them to get to the bottom of what
       may have happened.
   Q.  So there are actually reports you can use now?
   A.  There are reports written for those specific products.
   Q.  For those products --
   A.  Yes.
   Q.  -- that exist now --
   A.  Yes.
   Q.  -- and are being used?
   A.  As far as I'm aware, yes.
   Q.  Did you think it might be useful to have a look at
       those, or did you not look at them because of what you
       were told by the solicitors, that how it was done wasn't
       in scope?
   A.  Neither is accurate.  These reports were generated very
       recently.  As I say, this is a system that has been
       developed, it is in place, but the actual uptake by
       teams is being done on a step-by-step basis and as
       improvements are made, reports are written.  It's
       running quite well but organically.
   Q.  It is fair to say that your most recent witness
       statement is pretty recent as well, isn't it?
   A.  It is.
   Q.  8 March.
   A.  It is.
   Q.  So you could have captured information in that witness
       statement about the trends and --
   A.  That witness statement was to address specific comments
       within there that following understanding from looking
       at Mr Coyne's report, relooking at my statement, which
       is the right thing to do, it indicated that -- as per
       Dawn Phillips' testimony earlier that there was
       information in there that wasn't fully explained
       regarding notices.
   Q.  Two things, Mr Smith.  I know what's in the witness
       statement.  I was just pointing out that first of all
       it's not quite things that weren't fully explained.
       There was something in there that was wrong that you had
       to correct.  Can we be clear about that?
   A.  Terminology, I will agree, was wrong.
   Q.  It was wrong?
   A.  Yes.  The facts and the figures I would dispute were
       wrong.
   Q.  Right.  Well --
   MR JUSTICE FRASER:  Sorry, so you are saying the facts are
       correct and the figures are correct?
   A.  My Lord, as per Dawn's previous testimony she did say
       that the information that she provided, the information
       on error notices, the volumes that she provided, were
       accurate and that's all that I can work on.
   MR JUSTICE FRASER:  But if the description of what those
       figures is is incorrect then you are maintaining that
       the content of the statement is correct, are you?
   A.  The statement wasn't -- it wasn't correct, my Lord.  It
       was referring to error notices not transaction
       corrections.
   MR JUSTICE FRASER:  Right, Mr Green.
   MR GREEN:  I'm grateful.
           And in fact the heading in the second box, if we
       look at paragraph 23 on page 5 {E2/9/5}, the heading
       "Disputes received from branch" was specifically wrong,
       wasn't it?
   A.  It was, yes.
   Q.  Because it was actually "Disputes received by Santander
       from PO"?
   A.  It was, yes.
   Q.  So that the question I was originally asking you was you
       said that you didn't include anything about the
       transaction correction system or any analysis about TCs
       and so forth from the system because the reports were
       quite recent.  I showed you the date of your statement
       which was the Friday before this trial started and
       I suggested that was recent.  I'm just going to suggest
       again that would have been an opportunity for you to put
       something in about how those were working.  Was there
       a reason why you didn't?
   A.  There wasn't.
   Q.  Okay.  And there aren't any underlying documents
       produced or exhibited with your statement to support
       these tables, are there?
   A.  There isn't I'm afraid.
   Q.  So just to get a feel for how this was done, if we look
       at paragraph 17 which is {E2/9/4}, "Cash bureau and
       personal banking team", the people that you spoke to or
       contacted were Sarah Parkes and Gillian Hoyland, is that
       right?
   A.  It was.
   Q.  What did you do, did you email them, did you talk to
       them, did you call them, did you meet them; what
       happened?
   A.  There was an email that was quite explicit as to the
       information I required and I did do follow up emails and
       on occasion I would speak to people.
   Q.  On occasion you would speak to people?
   A.  Yes.
   Q.  Did you speak to Sarah Parkes?
   A.  I didn't speak to Sarah Parkes, no.
   Q.  Did you speak to Gillian Hoyland?
   A.  From memory, I did, yes.
   Q.  And which aspect did you speak to her about?
   A.  Just to request that the information was provided, to
       explain what the information was needed for and that it
       was time critical because I was still awaiting the data
       at the time.
   Q.  And you've got these average numbers.  Was there
       a reason why the period was different between the cash
       and the bureau figures?
   A.  That's not something I'm aware of.
   Q.  You don't know?
   A.  I don't know.
   Q.  And it suggests there that there are 680 disputes,
       paragraph 18, 40 upheld.  And you understand that the
       disputes tend to relate to alleged shortages in pouches
       received by branches and they are dealt with at the cash
       centre as opposed to FSC?
   A.  That's right.
   Q.  Who did you gain that understanding from?  Was that
       something that Gillian Hoyland told you?
   A.  That would have come from an email and I cannot remember
       who sent me the email at this stage.
   Q.  Okay.  Let's look at personal banking for a moment, if
       we may.  TCs by the personal banking team are rarely
       disputed, we see there.
   A.  Mm-hm.
   Q.  "This is because they run a report in Horace to
       determine whether a surplus is evident (ie if the cash
       physically counted is more than the amount that appears
       on Horizon or vice versa) before issuing a TC."
           Where has that come from?
   A.  That information is likely from Sarah Parkes but, as
       I have suggested earlier, I cannot 100% say which one of
       them provided that detail by email.
   Q.  Okay.  Camelot, debit card and ATM team, Andrea Green.
       {E2/9/5}:
           "The Camelot, debit card and ATM team had the
       following number of TCs disputed and issued the
       following number of compensating TCs in the 2017
       financial year."
           We have a table there.  Again, there is no
       underlying document provided?
   A.  No.
   Q.  How did Andrea Green give you this information?
   A.  Andrea Green will have sent this as an email which
       I will have forwarded on to our solicitors.
   Q.  And it just says -- under "BOI retracts" and "Lottery"
       the number of TCs issued for those two is very
       different, isn't it: 4,174 and lottery is 24,139?
   A.  It is.
   Q.  But the number of disputed TCs apparently is estimated
       to be the same?
   A.  It appears so, yes.
   Q.  And the number of compensating TCs is also estimated to
       be identical?
   A.  It appears so, yes.
   Q.  Do you have any feel at all from your own knowledge as
       to whether those figures are even remotely reliable?
   A.  The number of issued TCs I would suggest are very
       accurate.  The number of TCs disputed is not something
       that I can comment on at this stage.  I do know that
       prior to having our case management system in there was
       no consistent method of recording the method of
       disputes.
   Q.  You say in paragraph 21 what your colleague Andrea Green
       has told you.  So just working out the information
       flow --
   A.  Yes.
   Q.  -- it's you got it from Andrea and she got it from
       various experienced team leaders?
   A.  Team members.
   Q.  Team members.  Did she tell you who they were at all?
   A.  She didn't.
   Q.  Okay.  So it's hop over to Andrea Green and then another
       hop over to some other team members who are not
       identified?
   A.  Yes.
   MR JUSTICE FRASER:  That's really a point for me, Mr Green.
   MR GREEN:  I'm grateful, my Lord.
           So do you think there was scope for other
       misunderstandings between you and the people you have
       asked of the same sort as that which has emerged with
       Dawn Phillips?
   A.  It is unlikely.  I cannot say categorically no, but the
       contract and the way that we work with Santander is
       significantly different to how we work on the other
       products.  Many of the other products are very, very
       similar in how they are administered, there's just
       different methods of investigation and additional
       different types of data available.
   Q.  Okay.  Can we just go over the page please to page 6 in
       your witness statement {E2/9/6}.  DVLA:
           "Jacqueline has explained to me that TCs are very
       rare in this team.  During the past year the team have
       only issued four and none of them have been disputed."
           Now, you understand that this trial concerns the
       operation of the system over a long period?
   A.  Yes.
   Q.  Could we look at {F/1889} please.  Sorry, I hope I've
       got the right reference.  We can come back to that
       reference in a minute, if we can just check it.  It is
       the spreadsheet of TCs issued by branch.  We can come
       back to that in a second.
           If we can just go in the meanwhile to paragraph 30
       of your witness statement {E2/9/6} we see a reference to
       paragraph 6.66 of a report by Mr Coyne in which he
       refers to two TCs of £810,000 in a Post Office Limited
       disclosed document and suggests they might have been
       made in error.  You say:
           "This is not correct.  Two TCs were required in this
       instance because the method of payment used was cheque."
           Then:
           "TCs to the value of 810,00 are directly linked and
       relate to a Bank of Ireland Post Office Savings Account
       deposit on 24 August 2012."
           Let's quickly look at that.  It seems to be right.
       If we look at {F/1195} please.  You had a look at that
       Excel spreadsheet anyway to look at the two --
   A.  I did.
   Q.  There's an 810 and 810 at the top and the bottom, isn't
       there?
   A.  There is.
   Q.  And they cancel each other out.  What I wanted to ask
       you about is -- while that downloads -- it might
       download in the background if we go to another document.
       Shall we just look at {D3/1/205} please.  Let's get the
       spreadsheet, thank you very much.
           Can we go to the original data {F/1195} and there
       we've got a long list of individual TCs issued in 2012,
       with the date they are issued, going into 2013, the
       early part of 2013, so it looks like the 2012/2013-year,
       is that what you understood?
   A.  I haven't seen the full detail of this report, but it
       does appear that way.
   Q.  And if we go to the pivot table, the pivot report tab at
       the bottom we can see there are 84,217 transaction
       corrections recorded there, yes?
   A.  Yes.
   Q.  Could we please go to {D3/1/205}.  Now, this is
       Dr Worden's report, who is the expert for Post Office,
       and he has reproduced a table which was produced by
       Post Office in accordance with an order that the court
       made for provision of this information.  When we look at
       the volume of TCs in 2012 it says 124,374 and when we
       look in 2013 it says 105,876 and we can see that the
       moving average in that period is a lot higher than the
       figure that we saw in the financial year 2012/2013.  Are
       you aware of any reason why the numbers in what appear
       to be a calendar year might be very significantly
       different, other than the fact that the financial year
       starts and ends at a different time?
   A.  I know of no other difference other than it would be
       significantly different based on that premise.
   Q.  Can I ask you whether you have seen this document before
       please?
   A.  I don't recognise it, no.
   Q.  Have a look at page {H/58}.  This is the document as
       originally presented and sent through.  If we look at
       {H/57}, you can see what's said about this.  Just come
       down to the third paragraph:
           "Post Office does not hold historic aggregated TC
       data in a single searchable database, in Horizon or
       otherwise."
           We have seen that 2012 spreadsheet of all those
       transaction corrections individually and for what it's
       worth I have checked that there are 87,000 of them.
       Have you seen other spreadsheets of transaction
       corrections similar to that for different years?
   A.  Similar to that, no.
   Q.  The one we looked at which had the 810 and the 810 that
       you specifically looked at to check what Mr Coyne was
       saying?
   A.  Not in that layout, no, but Post Office do hold a --
       I will loosely call it a database, it's a SharePoint
       site, with Excel spreadsheets on that have been created
       to hold transaction corrections, which our -- one of the
       teams within Post Office appends to daily to create that
       list.
   Q.  Which department is in charge of that list?  Who is the
       owner of it?  I see it is on a SharePoint site, but who
       would be the natural person to talk about that list if
       they were asked?
   MR JUSTICE FRASER:  You have asked five questions, one after
       the other.
   MR GREEN:  I'm sorry.
   MR JUSTICE FRASER:  Do you want to maybe put it as one
       question.
   MR GREEN:  Who would be the natural person to speak to that
       list and explain it?
   A.  It will be somebody in the systems management team.
   Q.  Okay.  Can I ask you whether you had any involvement in
       or about May 2018 in looking at any of those sorts of
       documents to compile a table of how many transaction
       corrections there had been?
   A.  It's possible.  I cannot explicitly remember, but
       I could have.
   Q.  You can't remember?
   A.  No, not at this stage.
   Q.  Can we just look at {F/889}, which is my missed note.
       Can we look please at the "Branch type volume" tab
       please and you can see there's an analysis of
       transaction corrections by branch type, yes?
   A.  Yes.
   Q.  And then if we go to the "Summary by period" tab please.
       You can see there -- if we just take DVLA as an example,
       where your evidence of the present number of TCs is
       there is about a handful, if you go up to the top to see
       the headings there, can you see "2010/2011 outturn",
       that's an annual figure, that's 2,717, do you see that?
   A.  I do.
   Q.  So that's not very recognisable from what you have said
       in your witness statement.  Is that because that's gone
       down radically?
   A.  It will have gone down radically, yes.  I don't believe
       we actually transact that type of transaction with the
       DVLA any more.  The transactions that we're dealing with
       where there are very few is we now have a corporate
       relationship with the DVLA where we have fleet companies
       such as Arval or Hitachi Finance who tax their cars with
       us on a bulk basis and they pay us a significant amount
       of money maybe two days prior to when they are actually
       going to tax their cars and we request that the
       branches -- or we task the branches in calling us before
       they do these transactions to make sure that the funds
       are available to actually pay for these transactions.
   Q.  Okay.  Mr Smith, looking at the figures for DVLA in
       those earlier years that we have seen, it doesn't
       surprise you that they are of that order?
   A.  I don't have enough knowledge to know whether they are
       right or wrong I'm afraid.
   Q.  So you wouldn't have been able to assess very well what
       the other people you were getting information from were
       telling you?
   A.  No, I wouldn't.
   Q.  I'm grateful.
           My Lord, I have no further questions.
   MR JUSTICE FRASER:  Mr De Garr Robinson?
   MR DE GARR ROBINSON:  I have no further questions.
                 Questions from MR JUSTICE FRASER
   MR JUSTICE FRASER:  I have two.
           Can we go to your statement please, which is at
       {E2/9/3} and your paragraph 16 where you explain about
       this case management system.  Do you see the first line
       of your paragraph 16?
   A.  Yes, my Lord.
   MR JUSTICE FRASER:  What is it called?
   A.  It is called -- you will have to excuse me, nerves.  It
       is dynamics, my Lord.
   MR JUSTICE FRASER:  Dynamics.
   A.  Dynamics, my Lord, yes.
   MR JUSTICE FRASER:  I know this is quite new because you say
       it has not been fully rolled out and it only started
       in September 2018, is that right?
   A.  It did for the FSC, my Lord, yes.
   MR JUSTICE FRASER:  And then you also referred to reports
       which you said had been generated very recently.
   A.  Yes, my Lord.
   MR JUSTICE FRASER:  Do you recall Mr Green was asking you
       about those?
   A.  I do.
   MR JUSTICE FRASER:  What are they called?
   A.  I don't think they have a specific name, my Lord,
       I think they are just reports that have been set up
       within the Dynamics system to allow the team leader to
       easily access, such as a dashboard to be able to see
       graphically volumes and information behind that.
   MR JUSTICE FRASER:  Per month or -- because it has not yet
       been going a whole year -- or quarterly or what?
   A.  I believe it gives the data daily and monthly, but I'm
       going on a conversation as opposed to something I can
       give you concretely.
   MR JUSTICE FRASER:  That's all right.  It certainly can't be
       annually, can it?
   A.  No, absolutely not.  It could be set up to be annually,
       but it wouldn't give you -- it would just give you up to
       now, yes.
   MR JUSTICE FRASER:  At the moment it won't be an annual one.
   A.  Yes.
   MR JUSTICE FRASER:  That's very useful.
           Then the second question, if we can go forward
       please to {E2/9/6}, which will be your paragraph 27.  Do
       you see that?
   A.  Yes, my Lord.
   MR JUSTICE FRASER:  Now, the only reason I'm going to be
       a bit precise about this is because of the slight
       difficulty with headings in your tables earlier in your
       statement.  When you say "MoneyGram receive one or two
       disputes a month", is that MoneyGram the company
       receiving disputes from the Post Office, is that
       MoneyGram the team receiving disputes back from branches
       about TCs, or is it something else?  What does it mean?
   A.  My Lord, I believe this is MoneyGram as the team within
       the FSC.
   MR JUSTICE FRASER:  The MoneyGram team.
   A.  Yes, my Lord.
   MR JUSTICE FRASER:  Receiving one or two disputes a month
       from ..?
   A.  Branches.
   MR JUSTICE FRASER:  From the branches.  And is your
       understanding that they are disputes in respect of
       transaction corrections issued by the team, or in
       respect of something else?
   A.  My assumption based on what I put in paragraph 16, which
       is that this is the information that I have gained from
       team leaders, it would be disputes regarding transaction
       corrections that have been received by branch.
   MR JUSTICE FRASER:  But your paragraph 16 is talking about
       the sense of proportion of TCs that have been challenged
       and what challenges were successful.
   A.  I'm specifically --
   MR JUSTICE FRASER:  So the one or two disputes a month are
       back in respect of TCs that have been issued, is that
       right?
   A.  That is my assumption, yes, my Lord.
   MR JUSTICE FRASER:  And that would be as a proportion of
       what higher number, or any idea, or you can't remember,
       or you were never told?
   A.  I'm unaware of the number, my Lord.
   MR JUSTICE FRASER:  Unaware.  All right, that's very
       helpful, thank you.
           Mr De Garr Robinson, anything?  No.
   MR DE GARR ROBINSON:  My Lord, no questions.
   MR JUSTICE FRASER:  Mr Smith, that's your evidence done so
       thank you very much.  You don't have to come back
       tomorrow.  Obviously you are welcome to but you don't
       have to come back to give evidence.  Thank you very
       much.
           Anything that needs dealing with?
           Can I just raise a polite reminder about experts'
       reports cross-referenced to the Opus -- when I mentioned
       it last week it wasn't urgent.  It is still not urgent
       but I would broadly like them by the end of the week if
       possible.
           And then I was also going to have a file of PEAKs
       and KELs.
   MR GREEN:  My Lord, do you want us to give it to you as we
       go along?
   MR JUSTICE FRASER:  I think so, yes please.
   MR DE GARR ROBINSON:  It would be helpful if all parties
       could have the same file.
   MR JUSTICE FRASER:  I agree, but I don't suppose the
       claimants necessarily know which ones you are going to
       use when you are cross-examining Mr Coyne.
   MR DE GARR ROBINSON:  No, no, absolutely.
   MR JUSTICE FRASER:  I would have thought the task of
       providing one by the end of the week for the evidence of
       fact is probably going to be fairly straightforward,
       because we haven't been using very many.
   MR GREEN:  My Lord, yes.
   MR JUSTICE FRASER:  And then after that it can be updated
       every couple of days.  And it doesn't have to be updated
       before they are asked questions, it can be updated
       afterwards.  It's just to give me a useful working file.
           Right, anything else?
   MR DE GARR ROBINSON:  My Lord, no.
   MR JUSTICE FRASER:  No.  10.30 tomorrow.  Thank you very
       much.
   (4.36 pm)
        (The court adjourned until 10.30 am on Wednesday,
                          20 March 2019)