Price Waterhouse Coopers report on NLIS for Minister McGauran

The following five pages are key excerpts taken directly from:

"Report of the findings from a review of the operation of the National Livestock Identification Scheme"

by PricewaterhouseCoopers December 2006


The following are excerpts of the original text.
Edited by ABA Director John Carter.
Words have not been added or changed

AN MS-WORD VERSION OF THIS FILE WITH HIGHLIGHTS OF KEY POINTS CAN BE DOWNLOADED BY CLICKING ON THE LINK AT THE BOTTOM OF THIS PAGE

Preface-(Disclaimer)
The responsibility for determining the adequacy or otherwise of the procedures agreed to be performed by us was that of DAFF and the procedures performed were solely to assist DAFF’s assessment of the functioning of the NLIS system.

This report was prepared solely for the use of DAFF. PwC does not accept any responsibility to any other party to whom this report may be shown or into whose hands it may come.

The procedures that we performed do not constitute an audit in accordance with Australian Auditing Standards or a review in accordance with Australian Accounting Standards applicable to review engagements and, consequently, no assurance is expressed.

However, a number of issues were evident such that the NLIS system cannot be considered to be fully compliant with the NLIS business rules.

The majority of stakeholders also consider the NLIS system more efficient than the previous approach. – Unnamed – anonymous opinions.

In addition, most stakeholders believe that Australia would have been excluded from a number of major international markets had it not adopted a more effective animal tracing system. Unnamed – anonymous opinions. PwC did not investigate these matters, as it was not in their brief.

There are a range of issues that DAFF and operators of the NLIS system may wish to consider addressing.
The fact that notification of animal movements was not mandatory for a period is a large part of the cause of this. However, since notification became mandatory, there remains a significant rate (we were advised this represents approximately 6.4% of tags issued since the introduction of mandatory notification) of failure to notify the database of animal movements.

Animals that have been slaughtered not being recorded as such as a user-entry system (i.e. data is entered by an individual or their agent taking part in the transaction); there is the potential for erroneous data to be entered into the system through inadvertent or deliberate error. Such errors could be lessened, or more readily corrected, if there were regular user review of data entry. An enhanced mechanism, though at greater cost, would be to employ external scrutiny or review.

Perhaps the most important of these are the accounts of producers.-- Only around one in ten producers has opened an account and it is quite likely that a significantly smaller proportion regularly review their account. This lack of user or third party scrutiny means that many data errors are not discovered and are consequently never corrected.

A range of data transcription errors were evident during the review. These generally led to animals being listed on the wrong PIC, or having an incorrect deceased status. In some respects it may therefore be more appropriate to call the NLIS system a State based system on a national pattern rather than a single, national system. Differences in implementation in the various jurisdictions have created issues. These are particularly evident where animals that are subject to an exemption in one jurisdiction are moved to another.

Once full transaction recording is in place, theoretically a life record of an animal’s residency, and other animals it has interacted with, can be established. The intention is that this centrally stored, electronic history of an individual animal’s residency will enable rapid and accurate traceability of all cattle in Australia.

Trace a transaction through the end-to-end process to validate that the processes and controls documented are in place.

This discussion of the importance of livestock traceability does not presuppose that NLIS system is the most appropriate means of achieving that traceability.

Historical data (provided by the Victorian Department of Primary Industries) has shown that there is a major disease outbreak in Australia roughly once every four years. An outbreak such as BSE or Foot and Mouth, would likely result in total market exclusion for Australian livestock and associated products for a period of time until these products are considered disease free.

Stakeholders involved in the implementation of the NLIS system and in the disease and contamination issues of the 1990s and our review of the documentation associated with those incidents, indicated that Australia would have lost access to a number of significant international markets (principally the EU and Japan) had it not implemented a more effective livestock tracing system. Anonymous pushers of the NLIS.

In addition to the significant impact a disease outbreak would have on export markets, the number of cattle destroyed due to a major disease outbreak would affect the domestic market and industry participants including producers, stock agents, saleyards and abattoirs. Farmers would be forced to slaughter entire herds, costing the local industry many millions of dollars. --Anonymous supporters of the NLIS.

RULES AT ABATTOIRS
2.1 Within 48 hours of the relevant CVO2 being notified3, it must be possible to establish the location(s) 4 where a specified animal has been resident during its life.
2.2 Within 48 hours of the relevant CVO2 being notified3, it must be possible to establish a listing of all cattle that have lived on the same property as the specified animal at any stage during those animals’ lives.
Throughout 2005/06 over 10 million cattle movements were recorded, including over 4.4 million individual cattle movements from property of origin and almost 6 million cattle movements via sales through saleyards.
Only one jurisdiction met the section 2.1 criteria for all animals
The NLIS system provided a reliable, accurate tracing service, though allowance for older animals was needed in newly implemented databases
The NLIS system reports could be more user-friendly for jurisdictions
State access to the database would speed tracing rather than going through the database administrator.

AQIS AUDIT
Ten of the 40 animals, all without electronic ear tags, were never located. It must be noted that of these 10, 5 were moved interstate. However, it should be noted that the sample size of the audit was quite small. It should therefore be treated with some caution.

Some State legislation requires abattoirs to scan cattle prior to slaughtering so that information can be passed through the Easy Check software in order to determine the status of the beast. The abattoir will then know, prior to slaughter, whether cattle have a disease status, had hormone treatment or been exposed to chemicals.

In relation to scanned cattle, processors must up-load to the NLIS database the following information (by the close of business on the next working day or as otherwise required by State legislation) - abattoir establishment number, license number or PIC; NLIS device number; PIC of the last property of residence (this is not required for saleyard cattle in jurisdictions that require saleyards to report to the database the vendor PIC before slaughter cattle leave the saleyard); date of slaughter; and as a minimum Hot Standard Carcase Weight or weight at slaughter.

Historically however, there were problems in some businesses with this area of the technology. All businesses involved generally assured us that, even where scanners had failed to read tags, other processes were put into place to ensure that transactions were recorded. Anecdotally, however, there were suggestions that not all of the transactions were captured.

The actions required by individuals to enable the NLIS system to work effectively are set out above and in chapter 3. The consultation process revealed that not all of these requirements are always met.

The NVD captures information about the property in which the cattle have been including the PIC number and any treatment given to the animal. The NVD must be completed correctly in order to capture all the required information on cattle. If an NVD is not completed correctly, information required for traceability may be lost. However this issue does not specifically affect the operation of the NLIS database.
The contention has been put to us that the failure to electronically registered animal movements in these cases is not a problem because it simply means that authorities will need to fall back on the paper tracing system at some point. We do not agree with this view.
If the aim of the NLIS is to have an electronic, whole of life tracking system, then these failures to register transfers represent failures in the system.
Although tagging was compulsory, recording cattle movements on the database did not become mandatory until several years later (depending on which State the property was in). This meant that cattle which had been moved from the property of birth could lose their lifetime traceability because it is not be possible to determine where the cattle had been between the property of birth and the existing property unless a detailed paper trail existed.

Some States also introduced exemptions to the regulations such as that cattle sold directly to abattoirs did not require ear tags. These exemptions are generally no longer in place meaning all cattle must be tagged prior to leaving the property of birth. However, the Northern Territory still has an exemption for producer to producer transfers.
However, this does not affect the traceability of living cattle and will be rectified when the
database is archived.

It is largely this issue and the fact that MLA has not yet archived that has given rise to most, if not all, of the so called phantom cattle.

Many animal transfers have not been registered, even after it became compulsory to do so.

The reasons given for this have been varied. Most commonly it was believed that:

Property to property transfers were a major source of such problems because producers were unwilling to pay to have animals scanned, or were unaware of their responsibilities transfers between PICs owned by the same person were also seen as a common source of such problems
Technology failures at saleyards were often given as another cause.

It was stated a number of times that vendors generally had no responsibility or incentive to ensure that transfers were correctly registered and that purchasers, unless they were receiving some financial benefit from maintaining whole of life traceability, had little incentive to do so either.

However, this fall is not as large as we would expect. It tends to indicate that there are still a significant number of stakeholders who are not registering animal transfers.


These outcomes are obfuscated to some extent by the various exemptions that still exist in most states. For example, although the Queensland figure looks quite low, there has been an exemption in that jurisdiction for animals that proceed directly to slaughter - the majority of animals in that State.

In addition to this issue we have also seen at least some evidence of a few related issues including:

Producers clearing out their accounts by marking animals as deceased without verifying that they weren’t sold.
Animals being processed at abattoirs without being reported as deceased

Message to the Regulators - The States
With respect to the compliance of stakeholders with the State based legislative requirements of the NLIS system, we believe there are a number of issues which the Government and stakeholders may wish to address.
However, we had expected that the introduction of the mandatory requirement to inform the database of movements of animals from one PIC to another would have led to this problem being almost completely removed by this time. While mandating the requirement has reduced the incidence of the problem it has not removed it. At approximately 6.4% of animals sold (in October 2006) it is still an issue that needs addressing.

In addition we have seen at least some evidence that animals sent to slaughter have not always been notified as deceased. It is impossible to determine the extent of this issue but it is worth noting that it has occurred.
Apart from a few areas (particularly the system transfers effected) the database when an animal comes from an incorrect PIC) there is little third party or independent verification of the data entering the database.

The system largely relies on individuals - producers, saleyard and processor staff - entering the correct information. There is auditing of saleyards and abattoirs in most jurisdictions, however, the system largely relies on individuals doing the right thing and not making errors.

The incentives for many individuals to do the right thing are high where maintenance of whole of life traceability bestows a commercial benefit. However, where such a benefit is not immediately or personally evident, the incentive to fulfill all obligations and ensure the data is correct lessens.

Although not a requirement under regulations, some stakeholders have set up accounts within the database and use them to monitor their herd transfers and ensure that the correct cattle have been taken off their account and moved onto the purchaser’s. Checking accounts is an easy way to ensure that the system is being operated correctly and that the cattle remain traceable after leaving the property of birth.

However, most producers did not have accounts (less than ten percent have started an account) and most that did were not using it regularly. We found only one business (a major processor) that conducted regular audits of their account. As a result errors have crept into accounts and, where they are not addressed, can repeat, and therefore snowball, over time.

While this is not a failure in the internal system logic, it does mean that a potential method for finding and rectifying any errors that creep in is not being utilised. It also has the potential to undermine
confidence in the system.

The extent that an education program would be able to address this issue is limited by the proportion of people with PICs who have access to the internet and the technical ability to use such access to manage their accounts. A paper based account system (e.g. sending all PIC holders a quarterly or yearly statement of their account) would reduce this problem but would also come with attendant costs.

Most jurisdictions have undertaken only a handful of enforcement activities to date.

Enforcement is particularly important where there is little or insignificant commercial incentive to maintain lifetime traceability.

The NLIS system is not truly a national system in so far as it is not precisely the same in each jurisdiction. Rather it is a State level system on a national pattern

In addition to the cost of ear tags, some producers believe they are spending significantly more time inserting the NLIS tags than they did when using tail tags.

Although the majority of producers do not have active an account with the Database and do not transfer information regularly, producers in the past have not found the system user friendly, which has caused issues when transferring information.
The major issue associated with this is the additional cost incurred by the agent. Time associated with the scanning and recording of cattle information and administration time taken to send the information to the database costs agents. Some agents do not recovers or under recover this cost as it is difficult to accurately calculate the cost of the service. Other agents fully recover the costs of this service by charging the producers for all of the services they perform.

Several saleyards that purchased equipment that they later had to write-off when it did not work to a sufficient reliability or quality level stated their belief that the State Governments should bear some responsibility for such equipment.

Some abattoirs believed that identifying cattle in an abattoir is more difficult with ear tags than the old tail tag systems. The tail tag system allowed people to identify which region the cattle were from, which farm and whose cattle it was. possible, abattoir workers could visually identify each beast and know where it had come from.

Some producers believe that the Occupational Health and Safety risk has increased since the introduction of ear tags. These issues are mainly concerned with tagging full grown cattle without appropriate equipment such as a press. Many producers are now tagging livestock as calves to avoid the health and safety risk.

Our scope did not extend to undertaking any review of the general controls environment within which the database operates.

Prior to legislation introduced in January 2006, there was no mandated requirement for abattoirs to record animals as "deceased" into the database. This has resulted in legacy issue whereby a large number of tag devices recorded as active on the database for which no movements have been recorded or likely to be recorded.

We note that the ownership of the data lies with the various stakeholders, who are responsible for ensuring that the data they upload into the database accurately reflects the physical movements of cattle and other events. They are also responsible for ensuring that the data is free from error, including potential human error.

There is currently no disaster recovery plan (DRP) in place to recover the NLIS database and critical processes in the event of a serious incident or disruption such as one affecting the Telstra data centre, which hosts the MLA infrastructure.

We recommend that the steering committee formally assess the risks associated with a major disruption to the NLIS system, along with the cost or impact on operations. This in turn should guide the development of a suitable technological backup/infrastructure solution (if required) to deal with business requirements though the outsourced provider (Telstra).

END