WITNO9890500
WITNO9890500
Witness Name: Simon Recaldin
Statement No.: WITN09890500
Dated: 3rd September 2024
POST OFFICE HORIZON IT INQUIRY
Fifth Witness Statement of Simon Recaldin in the Post Office
Horizon IT Inquiry
1. I, Simon Recaldin, of 100 Wood Street, London, EC2V 7ER, say as follows:
Introduction
2. Except where I indicate to the contrary, the facts and matters contained in this
witness statement are within my own knowledge. Where any information is not
within my own personal knowledge, I have identified the source of my
information or the basis for my belief. In particular, my knowledge contained
within this statement is informed by Peters & Peters Solicitors LLP (“P&P”), who
are acting for Post Office Limited (“POL”) in respect of its Post Conviction
Disclosure Exercise (“PCDE”)' and criminal appeals and are advising and
supporting POL and its Inquiry solicitors, Burges Salmon and Field Fisher
(‘BSFf’) in respect of the Inquiry. I have been assisted in preparing this witness
statement by P&P. The facts in this witness statement are true to the best of my
knowledge and belief.
1 The PCDE included cases where (i) an individual was convicted of a criminal offence, (ii) POL (or Royal Mail
Group pre-separation) was the prosecutor, (iii) the prosecution was based wholly or partly on data derived
from Horizon, (iv) the prosecution was undertaken between 1999/2000 and 2013 (subject to a small number of
limited exceptions). The PCDE adopted an inclusive approach so that any cases that were not conclusively
outside of the criteria were included within it until verified as falling outside the criteria.
WITNO9890500
WITNO9890500
3. This witness statement has been prepared to update the Inquiry and core
participants in respect of further verification work that has been undertaken by
P&P in connection with historic POL prosecution and conviction data (‘the
further verification exercise”) since I gave oral evidence to the Inquiry on 29
September 2023. I summarise below the methodology adopted by P&P in the
further verification exercise, as a result of which I am now able to:
Update the number of convictions on the PFA list ie. the number of
Potential Future Appellants (“PFAs”)? to whom POL owed? a duty of
post-conviction disclosure);
Identify with more accuracy which prosecutions appear to have
related to conduct pre-dating the installation of Horizon;
Define with more precision the types of pre-Horizon prosecutions
with which it is most useful to compare the PFA list, in order to identify
whether, when and for what reasons the numbers of prosecutions
changed (see Appendix I);
Provide a more detailed breakdown of the number of prosecutions
and convictions per year prosecuted by POL that may have relied on
Horizon data (see Appendix II); and
Outline some potential conclusions about what the data now shows.
Why verification is an ongoing process
4. I previously explained‘ the source of historic data relating to POL prosecution
and conviction and mentioned that historically over 100
‘casework
spreadsheets were manually created and maintained by members of the POL
2 As referred to at paragraph 9 of my First Witness Statement dated 30 March 2023 and paragraph 33 of my
Second Witness Statement dated 13 July 2023.
3 I am advised that, as a matter of law, POL no longer owes a duty of post-conviction disclosure to anyone
whose conviction has been quashed by the Post Office (Horizon Systems) Offences Act 2024.
* paragraph 13 of my Second Witness Statement
WITNO9890500
WITNO9890500
Security Team during the period when prosecutions were taking place and
provided to (or in some cases identified by) P&P for the purposes of the PCDE.
5. During the PCDE, P&P identified many instances where the information in
casework spreadsheets was incomplete, incorrect or inconsistent, and took
steps to verify it, insofar as it related to cases potentially falling within the scope
of the PCDE, by cross-checking against other information, including underlying
case material and court records, where available and accessible by POL.
However, many gaps and inconsistencies still remain, especially for older cases
where POL has limited records. POL continues to receive information from
external sources, for example enquiries directly from individuals, or their family
members, who are able to provide POL with information of which it was
previously unaware such as the date, location and outcome of the
proceedings.® For these reasons, verification is an ongoing process and the
data continues to be subject to change as more information becomes available
and is only accurate as at the date of this statement. See also paragraph 20(i),
below.
The Further Verification Exercise
6. While POL’s historical prosecution and conviction data will never be fully
complete owing to the passage of time, gaps in its record-keeping and deletion
of records, it was decided that POL should take reasonable and proportionate
steps to ensure that its prosecution and conviction data is as up to date as
possible, including (where possible) in respect of cases falling outside the
scope of the PCDE.®
5 The number of these enquiries rose considerably after the broadcast of Mr Bates v The Post Office in early
January 2024. By way of example, two individuals who appeared on the casework spreadsheets in 1998 and
2001 respectively, but for whom no prosecutor or case outcome had been recorded, provided POL with
documentation demonstrating that they had been prosecuted and convicted by POL. Prior to receiving this
information, POL had not been able to find any records in relation to either individual that confirmed the
conviction or which agency conducted the prosecution.
° In particular, data relating to cases from the 1990s and cases which did not result in a conviction are often
unverifiable owing to data retention policies and/or the manner in which the data was /is held. Also, as the
PCDE was concerned with convictions arising from the use of Horizon in which POL owed a duty of post-
WITNO9890500
WITNO9890500
7. Accordingly, in October 2023, POL asked P&P to commence a further
verification exercise to look at historic POL prosecution and conviction data
including in relation to cases outside the scope of the PCDE, so that POL could
provide more complete, reliable and consistent answers to questions about
investigation, prosecution and conviction data both before and after the rollout
of Horizon, with a view to updating the POHIT Inquiry and responding to
Freedom of Information Act (FOIA) requests from members of the public.
8. P&P inform me that the verification exercise included:
a. collating all casework spreadsheets maintained by the POL Security
and Legal teams which could be identified following the collection of
data for the PCDE.
b. creating a ‘Master Casework Spreadsheet’ into which all entries from
those casework spreadsheets were transferred, with a consolidated
entry created for each case reference.’ This also included
prosecutions which did not result in a conviction.
c. adding further data relating to cases about which POL had become
aware from external sources.
d. Where the quality of the data allowed (given that limited and/or
ambiguous and/or contradictory data is held for a significant number
of cases), making a final determination as to the outcome of the case
and other key matters such as the identity of the prosecutor and
conviction disclosure, to date it has not been necessary to verify pre-Horizon and non-conviction data to the
same extent as data relating to cases falling within the scope of the PCDE.
7 Many cases appear multiple times across different casework spreadsheets. The data in each casework
spreadsheet has been deconflicted and verified as far as possible and then consolidated into a single entry on
the Master Casework Spreadsheet.
WITNO9890500
WITNO9890500
whether the investigation related to conduct occurring before or after
the installation of Horizon in branch.
e. adding further information or corrections to the information recorded
regarding Branch Names, FAD codes, names, courts, conviction
dates etc.
f. comparing some information deriving from the casework
spreadsheets with the underlying material, where available, to
highlight any discrepancies, which were then investigated and
resolved.
g. determining an approximate ‘case creation date’, if none was
recorded on the casework spreadsheets, to enable cases to be
categorised by the date or year in which the criminal investigation
commenced.®
h. identifying from the ‘case types® indicated on the casework
spreadsheets those cases that are considered to be:
(i) Within the scope of the PCDE or, for pre-Horizon cases, are
equivalent to, or ‘like for like’, cases that are within the scope of
the PCDE, i.e. had those cases been prosecuted post-Horizon,
they would have necessarily relied in whole or in part on Horizon
data as if Horizon had already been installed (collectively
“Horizon-related” cases); and
5 I am informed by P&P that data showing the date of the alleged criminal conduct is not routinely available on
the casework spreadsheets. However, for the majority of entries a ‘case creation date’ is recorded showing the
date on which the criminal investigation commenced. Therefore, in the absence of underlying records showing
the date of the alleged criminal conduct, the most accurate indicator of whether the alleged criminal conduct
relied on Horizon data is to cross-refer the case creation date with the date Horizon was installed in the branch
in question (either permanently or ona trial basis). Where the case creation date is earlier than the Horizon
installation date, the case is considered to be pre-Horizon and therefore outside the scope of the PCDE.
° These are the case types recorded on the casework spreadsheets by POL at the time. There are dozens of
such case types, many of which appear to overlap, and it is not always clear precisely how the case type was
chosen or how each one was defined.
WITNO9890500
WITNO9890500
(ii) ‘Shortfall-related’ meaning a sub-set of the Horizon-related
cases which involved a shortfall or accounting discrepancy.
i. These sets of case types could then be used to compare the numbers
of prosecutions and convictions pre-Horizon for the same kind of
conduct and offences that were prosecuted post-Horizon’s
installation. A table showing the lists of case types recorded on the
casework spreadsheets, and those which P&P considers to be
Horizon-related and shortfall-related is at Appendix I.
9. The verification exercise was completed in May 2024. The product of this
exercise is a ‘Master Casework Spreadsheet of Investigations, Prosecutions
and Convictions’ ("MCS"). While there may be more work that could be done to
fillin gaps in POL’s records (particularly in respect of cases which did not result
in a conviction), the approach that has been taken is considered to be
reasonable and proportionate.
10.As explained above, the MCS will continue to be updated as and when new
information becomes available. Therefore, POL’s assessment of the numbers
of prosecutions and convictions in both the pre-Horizon and post-Horizon
periods is as accurate as it reasonably can be in the circumstances, as at the
date of this statement.
Results of the verification exercise
11.During my oral evidence to the Inquiry, I was asked whether there was a
significant increase in prosecutions following the implementation of Horizon and
/ or shortly before. I explained that there was an increase in convictions coming
up to the introduction of Horizon and there was a continued increase in
convictions after the introduction of Horizon.
WITNO9890500
WITNO9890500
12.Generally speaking, POL has less data for cases prior to 2003, therefore such
a comparison is not straightforward. Specifically, in attempting to respond to
various FOIA requests asking for comparative pre and post-Horizon data ‘° it
became clear that, beyond looking at total numbers for all case types (which
would include e.g. robberies and burglaries with no apparent connection to
Horizon), it was not easy to compare the number of Horizon-related cases
(post-Horizon installation) with the number of ‘like for like’ cases prosecuted
pre-Horizon installation.
13.At Appendix II, I include a table created by P&P that is based on the data in the
MCS containing information currently known to POL regarding prosecution and
conviction data’' from 1990 to 2020. This is broken down by:
a. All cases;
b. Horizon-related cases (i.e. cases within scope of the PCDE and all ‘like
for like’ pre-Horizon cases as defined in Appendix I); and
c. Shortfall-related cases (i.e. all cases of the types defined as ‘shortfall-
related’ in Appendix I).
Use of case creation dates
14.The table at Appendix II distinguishes between pre and post-Horizon cases by
comparing the case creation date with the Horizon installation date as explained
at footnote 8. Where the case creation date is prior to the Horizon installation
date, the case is considered to be pre-Horizon.
1° As I explained in my Second Witness Statement, POL has received a number of FOIA requests with different
parameters which produced different results depending on the date range and, for example, on whether they
asked about numbers of prosecutions or only convictions; all types of defendants or only sub-postmasters etc.
4 (This data is limited to cases prosecuted by POL in England and Wales and does not include cases prosecuted
by other prosecutors or those in Scotland and Northern Ireland.
WITNO9890500
WITNO9890500
15.The totals for each year are based on the case creation date rather than the
date of conviction. In my view, this allows for a more accurate analysis of the
potential impact of the introduction of Horizon on the number of prosecutions
and convictions, because it more accurately identifies whether the alleged
offence occurred before or after Horizon was installed at the branch in question.
Using the year of conviction is unhelpful for such an analysis, as there are many
cases in which the alleged offence occurred before Horizon installation, but the
conviction was after installation. It also means that cases with unknown
conviction dates can be included in the analysis because these generally have
case creation dates (and where they do not, approximate case creation dates
have been used).
Inclusion of additional case types to enable a like for like comparison
16. Further consideration has been given to the categorisation of cases by the type
of alleged conduct. At the time of my Second Witness Statement, figures for
shortfall-related cases were provided in brackets in Appendix II of my Second
Witness Statement for the pre-2000 period (and in notes for the years 2000,
2001 and 2002) in order to facilitate comparisons between Horizon-related
cases and similar cases pre-2000.'2
17.As explained at footnote 45 of my Second Witness Statement, the categories
considered likely to involve a shortfall in branch accounts included cases
described on the casework spreadsheets as theft, false accounting, audit
shortage and cash loss. However, a significant proportion of cases included in
the PCDE, including some overturned on appeal, involved pension and
allowance (“P&A”) frauds*’, which were excluded from the definition of shortfall-
related cases in the pre-2000 period because such cases are unlikely to involve
a shortfall in the accounts.
2 I explained at paragraph 11 of my Second Witness Statement that “I understand that the pre-2000 figures
and post-1999 figures are not directly comparable...”
33 A P&A fraud typically involved the manipulation of benefit payment cheques (e.g. suppression, submitting
them twice, cashing cancelled cheques etc) to create a surplus in the branch accounts which the defendant
then took for themselves, leaving the account balanced. Therefore, a typical P&A fraud did not involve a
shortfall in the branch accounts.
WITNO9890500
WITNO9890500
18.POL has always included all cases within the PCDE that may have relied, in
whole or in part, on Horizon evidence. This covers a much wider set of case
types than only shortfall-related cases, for example, P&A frauds. Consequently,
Appendix II to this statement shows all Horizon-related cases both pre and post-
Horizon installation, which allows for a more accurate and transparent
comparison between the numbers of post-Horizon cases, and pre-Horizon
cases that would have relied on Horizon evidence had they been prosecuted in
the post-Horizon period. I have, however, also set out the details of the shortfall-
related cases both pre and post-Horizon, to demonstrate the cases within this
time period that related to a shortfall or accounting discrepancy.
19.Unlike Appendix II to my Second Witness Statement, a breakdown of cases by
the role of the individual prosecuted has not been included in the table as this
is considered to be of less relevance for the purposes of identifying the potential
impact of the introduction of Horizon on the number of prosecutions and
convictions.
Caveats to the data
20.The data in these tables should continue to be treated with caution for the
following reasons:
(i) More information may come to light. Whilst a significant amount of work
has been undertaken to identify all investigation, prosecution and
conviction records for inclusion in the MCS, the data remains incomplete,
with known gaps."4 As stated above, POL continues to receive requests
for disclosure from individuals for whom it has no record of any
prosecution. Some of these individuals are not referenced at all on the
MCS (and therefore any of the casework spreadsheets), and work is
14 For example, the case reference numbers for very early cases follow on sequentially from one another (e.g.
Case References 1, 2, 3, 4, 5 etc). However, there are gaps in these sequences, such that it suggests there were
cases for those reference numbers, but POL no longer holds data for those cases.
WITNO9890500
WITNO9890500
being undertaken to ascertain who prosecuted those cases and their
outcomes.
(ii) There are some cases for whom there is a record on the MCS, but the
record indicates that there was no recorded outcome, whereas
information received from the individual has confirmed both that POL
prosecuted the case and that a conviction was secured.
(iii) I The data in the appendices relates only to cases prosecuted by POL (or
Royal Mail Group pre-separation) and cases where it has not been
possible to confirm the identity of the prosecutor, and not to cases
confirmed as prosecuted by other agencies.">
(iv) There are a limited number of casework spreadsheets capturing pre-
2003 data. For example, according to the available spreadsheets there
appear to be only two convictions secured in the 1980s, which is
considered to be unlikely. The data deriving from the casework
spreadsheets should not be considered an accurate record of the real
number of convictions secured in the 1980s and 1990s, particularly as
there is very little available underlying data to confirm the accuracy of
the spreadsheets for this period.
(v) For some cases, no case creation date is recorded on the casework
spreadsheets. To avoid having to record all of these cases in an
‘unknown’ date category, which would exclude them from the analysis,
P&P have estimated the likely year or decade in which the investigation
commenced based on other information contained within the casework
spreadsheets (i.e. by looking at the case creation dates of cases opened
around the same time, based on their reference numbers). This
approach was considered preferable to the approach taken to Appendix
+8 POL's casework spreadsheets also include cases prosecuted by the Crown Prosecution Service, Crown Office
and Procurator Fiscal Service, Public Prosecution Service of Northern Ireland, Department of Work and
Pensions, and Royal Mail Group (post-separation).
10
WITNO9890500
WITNO9890500
llin my Second Witness Statement, which included a significant number
of cases in an ‘unknown’ date row.
(vi) It is acknowledged that relying primarily on case types to identify ‘like for
like’ pre-Horizon cases (i.e. that would have relied on Horizon evidence
if prosecuted post-Horizon) is an imperfect exercise. However, owing to
the absence of underlying records, particularly in early cases, in the
circumstances I consider this to be the best available method.
21.It is also recognised that the criteria adopted for the PCDE are different from
the conditions for a “relevant offence” in s.2 of the Post Office (Horizon
Systems) Offences Act 2024, which defines the convictions that are now
quashed under s.1 of the Act. Therefore, the number of convictions on the PFA
list may well differ from the number of POL convictions quashed by legislation.
Updated number on the PFA list
22.During the further verification exercise information has been identified which
has resulted in a change to the number of individuals previously falling within
the scope of the PCDE. Overall, the number of PFAs has decreased from 700
at the time of my Second Witness Statement on 12 July 2023 (it remained
unchanged at the time I gave oral evidence to the Inquiry on 29 September
2023), to 697 as at today’s date. Although the overall number has reduced,
there are a number of cases that have been added to the PCDE as well as a
number that have been removed. Cases have been removed for one or more
of the following reasons:
a. The case has been confirmed as having been prosecuted by another
prosecutor.
b. Horizon has been confirmed by available records as having been
installed after the investigation was commenced, indicating that the
alleged offences must have occurred prior to the installation of Horizon
in the branch in question.
11
WITNO9890500
WITNO9890500
c. The prosecution took place within the PCDE period, but the case did not
rely in whole or in part on Horizon data (e.g. theft of mobile phone
vouchers).
d. The outcomes of a number of cases that were thought to have been
potential convictions have now been identified, and these have been
confirmed as not having resulted in a conviction.
23.Cases have been added for one or more of the following reasons:
a. Aco-defendant’s conviction was not previously identified largely owing
to the way in which multiple defendant prosecutions were recorded on
the casework spreadsheets.
b. Horizon has been confirmed by available records as having been
installed before the investigation was commenced, indicating that
Horizon evidence may have been relied upon in the original
prosecution. 1®
c. The case was previously thought to be a non-Horizon case, but following
a review of the underlying paperwork, it cannot be excluded that the case
may have relied on data deriving from Horizon.
d. Information contained in the comments of the casework spreadsheets
suggests that the person was convicted. Whilst it is likely that these
cases were not prosecuted by POL (given the way in which the data has
been captured for these cases, which is not how POL would capture data
regarding its own prosecutions), as it cannot be confirmed that they were
prosecuted by another prosecutor, POL has counted them as having
been prosecuted by POL.
1© I understand that these are thought likely to be DWP prosecutions but there is no underlying material
available to confirm the prosecutor and whether or not they are Horizon-related so they have been treated as
possible POL prosecutions that may have relied on Horizon and therefore included on the PFA list.
12
WITNO9890500
WITNO9890500
Conclusions that can be drawn from the data
24.At paragraph 72 of my Second Witness Statement, I concluded that, “/t appears
from the data at Appendix II, that the conviction figures increased in the years
shortly before the roll-out of Horizon (per the figures included in the notes in
Appendix II), with further increases after the roll-out. However, I understand that
this data should continue to be treated with caution given the limitations of the
pre-2000 data which has not been verified for the purposes of the PCDE or the
Inquiry, and given the pre-2000 data and post-1999 data is not directly
comparable.”
25.Subject to the caveats at paragraph 20 above, I believe it might be possible to
draw a number of conclusions from the data in Appendix II, as illustrated in the
following charts:'”
Figure 1: Post Office prosecutions 1990-2015
120
100
80
60
40
20
0
FIFI I EES ILLS SILI EFSF PPP PP Py?
All prosecutions + ——'Like for like’ prosecutions <=—Shortfall-related prosecutions
1’ Please note that all charts only include cases in definitive years, rather than those in the “1990s” or “2000s”
category where the exact year of case creation is unknown.
13
120
100
80
60
40
20
120
100
2
i)
FS
3S
Ny
8
0
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Figure 2: 'Like for like' and shortfall-related convictions per
year
1996
1997
198
1999
2000
2001
2002
3
200
2005
2006
207
ie
200° a
nn
=
212
2013 1
1990 I
1991 IE
1992 I
1993 B
1994
1995 &
2014 I
2015
mShortfall-related prosecutions Like for like’ prosecutions
Figure 3: 'Like for like' prosecutions pre and post Horizon
PP Mri gh
™Pre-Horizon mi Post-Horizon
Figure 4: Shortfall-related prosecutions as a % of ‘like for like’
prosecutions
PIII I ELEI FEES SESS LIS ESP P PP Go?
WITNO9890500
WITNO9890500
14
WITNO9890500
WITNO9890500
26.Figure 1 shows the total number of prosecutions per year for all case types,
‘like for like’ cases and shortfall related cases. Figure 2 shows the total number
of ‘like for like’ prosecutions per year and how many of them were shortfall-
related. Figure 3 shows the total number of ‘like for like’ prosecutions per year,
broken down by whether they were pre-Horizon or post-Horizon.'® Figure 4
shows how the proportion of ‘like for like’ cases that were shortfall-related
changed over time.
27.From these charts, I believe it is possible to conclude that:
a. The number of prosecutions brought by POL for ‘like for like’ offences
(i.e. offences which would have relied on Horizon data in the post-
Horizon period) began to rise in the mid-1990s, prior to the rollout of
Horizon.
b. The number of prosecutions stayed, on average, at this higher rate
following the rollout of Horizon until prosecutions ceased.
c. Although the rollout of Horizon began in 1999, all prosecutions in that
year and most of the prosecutions in 2000 were for alleged offences
committed before the installation of Horizon in those branches. From
2001 onwards, all ‘like for like’ prosecutions were for alleged offences
committed after the installation of Horizon.
d. There was no clear and obvious ‘spike’ (whether in all prosecutions,
Horizon-related prosecutions or shortfall-related prosecutions)
attributable to the rollout of Horizon.
e. The proportion of prosecutions which were shortfall-related began to rise
after the rollout of Horizon, including a significant rise from 2006
onwards. From that point onwards, 82% of prosecutions brought by POL
were for a shortfall-related offence.
18 As set out at paragraphs 14 and 15, a pre-Horizon case is one in which, according to the case creation date,
the investigation began, and therefore the alleged offences had already been committed, prior to the
installation of Horizon in branch.
45
WITNO9890500
WITNO9890500
28.In my view it is not possible to draw any definitive conclusions from the data as
to why the numbers and types of prosecutions changed over time. There
appears to be a clear correlation between the use of Horizon and an increase
in the number and proportion of shortfall-related cases, although this increase
does not appear to have been immediate.
Pre-Horizon use of ‘Capture’
29.POL is making enquiries into the claims by postmasters, MPs and in the media
that the computer system known as ‘Capture’, which postmasters could
purchase from POL’9, may have been responsible for miscarriages of justice.
I understand that Capture was first made available in 1992 and appears to have
been used by around 10% of branches”° until Horizon was introduced. POL
has not been able to locate a list of all branches which used Capture and it is
not possible to identify from POL’s data the numbers of prosecutions which may
have relied on evidence from Capture, but I am aware that some pre-Horizon
prosecutions were for alleged offences committed in branches which used
Capture.
Statement of truth
I believe the content of this statement to be true
Signed::
Dated: 3" September 2024
19 Capture consisted of a standalone computer and floppy disk that allowed the user to input all transactions
manually rather than having to hand-write them on paper forms. Once the postmaster had inputted the
transactions, they could use the software to produce balances, cash accounts and lists of pension and
allowance payments for submission to the DWP.
® Current investigations by POL have identified 488 Capture installations in October 1992, when the system
was first rolled out, followed by 1,100 Capture installations in January 1994, rising to 1,400 installations by July
1994. In 1998, we understand a further 2,500 copies of Capture were produced for an updated version of the
system known as ‘C90’. At the time of launch in 1992 there were circa 18,000 Post Offices, indicating that
around 10% of Post Office branches nationally used the Capture software.
16
Appendix I - Case type determinations in prosecution and conviction cases
WITNO9890500
WITNO9890500
Case type recorded
on the MCS
Horizon-related (i.e.
within scope of the
PCDE or treated as
‘like for like’ pre-
Shortfall-related
Horizon)
1. I Allegations affecting Yes Yes
or against
character?"
2. __I ATM (Inflated) Yes Yes
3. I Audit shortage? Yes Yes
4. Banking Fraud Yes Yes
5. Banking Post Office Yes Yes
Card Account Fraud”?
6. Banking Credit / Debit Yes Yes
Card Fraud’
7. I Bureau Theft/Fraud” Yes Yes
8. _I Business Concern’ Yes Yes
9. I Cash Loss Yes Yes
10. I Cash Loss (Audit Yes Yes
Discrepancy)
11. I Cheque (inflated) or Yes Yes
Cheque Fraud
12. I Credit/Debit Card Yes Yes
Fraud
13. I Crown Office or Yes Yes
Directly Managed
Branch Cash Loss?”
14. I Disputed Transaction Yes Yes
(s)8
15. I False Accounting Yes Yes
16. I Fraud (Internal)? Yes Yes
17. I Lottery Yes Yes
21 There are a number of anomalies in this Case Type, as allegations affecting the character sometimes related
to non-financial related crime such as possession of indecent images; blackmail etc.
2 Or “Audit” or “Audit / False Accounting”.
?3 Unless it is clear from the data that it was not a ‘like for like’ or shortfall related case.
4 Unless it is clear from the data that it was not a ‘like for like’ or shortfall related case.
25 Or “Bureau” or “Bureau Loss / Fraud”.
?6 Unless it is clear from the data that it was not a ‘like for like’ or shortfall related case.
77 Unless it is clear from the data that it was not a ‘like for like’ or shortfall related case.
28 Unless it is clear from the data that it was not a ‘like for like’ or shortfall related case.
2° Unless it is clear from the data that it was not a ‘like for like’ or shortfall related case.
A?
WITNO9890500
WITNO9890500
Case type recorded
Horizon-related (i.e.
Shortfall-related
on the MCS within scope of the
PCDE or treated as
‘like for like’ pre-
Horizon)
18. I Mint Redemption Yes Yes
Fraud
19. I Missing Voucher Yes Yes
20. I Other’? Yes Yes
21. I Post Office Card Yes Yes
Account*"
22. I Stock (Inflated) or Yes Yes
Stock Loss
23. I Theft Yes Yes
24. I Theft of Official Cash Yes Yes
25. I Theft / False Yes Yes
Accounting
26. I DVLA Yes No
27. I DWP Overclaims / Yes No
Fraud / IR Girocheque
Exception Fraud
28. I Giro suppression Yes No
29. I P&A Encashment Yes No
Fraud
30. I P&A Laundering Yes No
31. I P&A Overclaims Yes No
32. I P&A Fraud Yes No
33. I Postal Order Fraud / Yes No
Theft
34. I Theft (P&A) Yes No
35. I Suppression / Yes No
Suppression of
Banking Document
36. I Armed Robbery No No
3° Unless it is clear from the data that it was not a ‘like for like’ or shortfall related case.
31 Unless it is clear from the data that it was not a ‘like for like’ or shortfall related case.
18
WITNO9890500
WITNO9890500
Case type recorded
Horizon-related (i.e.
Shortfall-related
on the MCS within scope of the
PCDE or treated as
‘like for like’ pre-
Horizon)
37. I Attempted Armed No No
Robbery
38. I Burglary No No
39. I Cash Services / No No
Supply Chain —
Remittance / Plastic
Bank Note Envelope
Dispute*?
40. I Cash Services / No No
Supply Chain — Cash
Centre Internal Cash
Loss**
41. I Compensation Fraud No No
42. I Conspiracy to Defraud No No
/ Obtaining by
Deception™
43. I Counterfeit Goods No No
44. I Counter Snatch No No
45. I Deception No No
46. I LBU Discrepancy®® No No
47. I Motor Vehicle Licence No No
Loss
48. I Mail Offence (Theft / No No
Tampering / Delay)
49. I Misuse of Postage No No
Labels
50. I Non Branch Related No No
51. I Obtaining Postal No No
Services by Deception
52. I Offences against the No No
person
53. I PBNE Discrepancies*® No No
54. I Remittance Shortage*” No No
55. I Robbery No No
» Unless branch related.
33 Unless branch related.
4 Unless it is clear that itis a like for like or shortfall related case.
35 Relates to non-branch related losses.
3° Unless it is clear from the data that it possibly is a ‘like for like’ or shortfall related case.
37 Unless branch related.
19
WITNO9890500
WITNO9890500
Case type recorded
Horizon-related (i.e.
Shortfall-related
on the MCS within scope of the
PCDE or treated as
‘like for like’ pre-
Horizon)
56. I Suppression No No
57. I Theft of postal packets No No
from over counter
58. I TV Licence Fraud / No No
Theft of Postal Orders
20
WITNO9890500
WITNo9890500
Appendix Il — Post Office prosecutions and convictions 1990 — 2020
Pre-Horizon Post-Horizon Total Pre-Horizon Post-Horizon Total Pre-Horizon Post-Horizon. Total
Pros. I Conv. Pros. I Conv. Pros. Conv. Pros. I Conv. Pros. I Conv. I Pros. I Conv. Pros. I Conv. I Pros. I Conv. I Pros. I Conv.
1990 1 1 1 1 1 1 1 1 oO 0 0 0
1991 6 5 6 5 4 3 4 3 3 2 3 2
1992 2 0 2 0 2 0 2 0 0 0 0 0
1993 T 4 7 4 6 3 6 3 2 1 2 1
1994 10 9 10 9 9 8 9 8 4 3 4 3
1995 T 5 7 5 7 5 7 5 0 0 0 0
7996 31 27 31 27 26 23 26 23 15 14 15 14
1997 60 54 60 54 39 36 39 36 18 16 18 16
1998 93 86 93 86 70 65 70 65 29 29 29 29
1999 112 105 z a 114 107 101 94 101 94 22 20 22 20
2000 64 56 39 36 103 92 59 51 39 36 98 87 20 19 10 10 30 29
2001 69 61 69 61 63 56 63 56 28 24 28 24
2002 78 65 78 65 75 62 75 62 31 28 31 28
2003 68 51 68 51 58 47 58 47 34 27 34 27
2004 104 98 104 98 94 90 94 90 34 32 34 32
2005 73 65 73 65 64 57 64 57 44 39 44 39
2006 60 51 60 51 48 45 48 45 43 40 4a 40
2007 52 47 52 47 51 46 El 46 35 32 35 32
2008 7 76 7 76 7 76 7 76 70 69 70 69
2009 64 61 64 61 63 60 63 60 59 ST 59 57
2010 41 37 41 37 41 37 41 37 36 34 36 34
2011 44 a1 44 a1 a1 39 41 39 32 32 32 32
2012 44 at 44 at 44 at 44 at 38 35 38 35
2013 3 3 3 3 3 3 3 3 3 3 3 3
2014 2 2 2 2 2 2 2 2 2 2 2 2
2015 1 1 1 1 0 0 0 0 0 0 0 0
2016
2017
2018 1 1 1 1 i} ) 0 0 ) C) ) i}
2019
2020
7990s® I 40" 7 70" 7 Po on Oo om rr ro ro a
2000s” 3 2 3 2 cad 0 a 0 0 7 0
Total 403 359 825 741 1228 1100 333 295 764 697 1097 992 17 106 500 464 617 570
3 Unknown Case Creation Date but understood from all available data to be in the 1990s.
® Unknown Case Creation Date but understood from all available data to be in the 2000s.
Entries with an asterisk relate to cases that occurred in branches in which the Horizon pilot was installed. However, these cases are convictions of non-branch staff (i.e. outsiders) in relation to robberies and counter snatches.
"Entries with a double asterisk include cases where the case type is unknown, and therefore is included as a like for like/shortfall case out of an abundance of caution.
21