Independent analyst Simon Quilty takes a deep dive into the reasons behind the recent four million head adjustment to ABS’s national herd size assessment, and what it means for industry. Beef Central first wrote about the change in this article last Friday. “The change in the size of the herd is, for some, a necessary evil. We need to get behind these new figures, support ABS and rebuild market confidence in the data,” Mr Quilty concludes. “The challenge is explaining this dramatic change to customers worldwide, who may not understand that nothing has changed regarding supply from Australia.”

 

 

 

LAST week, the Australian Bureau of Statistics released a revised herd size estimate for Australia, adding 4.34 million head to the June 2023 estimate of 29.88m.

So, have four million head of cattle appeared magically in the system, and can we expect large volumes of cattle to come forward?

Industry Concerns: this revision has sparked numerous concerns among industry participants, particularly in light of the downward trend in slaughterings over the past decade and the anticipation of a record-low slaughter year next year.

In short, these figures contradict our beliefs. While these figures would imply near historical highs of the herd size, today’s kill and the last ten years imply the opposite.

I spoke at length last week with Rob Walter, Director of the Agriculture Statistics Program at ABS, who answered many of my questions about the changes made with the new methodology. This week, I followed up with the ABS analytical team on the additional concerns raised in this paper.

This paper explains why ABS has done this, outlines the new methodology compared to the old, and discusses the pluses and minuses of this new set of numbers and what it potentially means going forward.

In brief, I am looking forward to answering the following questions.

• Why has ABS changed its number drastically, and what is the new methodology?
• What is the risk to the market of such a dramatic increase in cattle numbers?
• What are the concerns with this new methodology? Will it result in more accurate or less accurate figures in the future?
• What are the positives of this new methodology?
• Does this methodology compromise ABS independence?
• What does this mean going forward?

Why has ABS changed its number drastically, and what is the new methodology?

Due to the government’s lack of funding and a perceived low response rate to their last 2022 herd and flock size survey, the ABS has developed a new methodology that uses an array of data sources to estimate herd size.

Outline of new methodology versus the old

The old method was a snapshot of Australian herd size by each state and nationally based on survey responses. The survey was done yearly, with almost 25,000 agricultural operations approaching, and every fifth year at the time of census, this survey pool was increased to 100,000 agricultural operations.

This separated beef and dairy herds and broke the herd into categories based on calf numbers, cow and heifer over one year, and the ‘other’ category.

The last survey conducted in 2022, had a response rate of close to 60pc, or 15,000 agricultural operations. This is not the first time that response rates have been low, and when this has occurred in the past, the ABS team has looked to other data sources to understand the trends better.

This was the last survey conducted, and now ABS has moved to a multi-data collection process based on flows into and out of the herd.

The new methodology for estimating cattle herd and sheep flock numbers uses fertility data, movements, and slaughter on a regional basis across Australia; included are animal disposals, live exports, and information on herd/flock demographics and fertility, together with rainfall and pasture conditions.

The 2023 cattle herd number has been released as an experimental series, and the ABS states that this ‘reflects that there will be ongoing development to further refine the method in addition to applying it to the sheep flock’.

It should be noted that there is no longer any ‘Estimated Value of Agricultural Output (EVAO) of $40,000 per property’, as the new methodology is based on livestock flows and the definition of all cattle grown for a commercial purpose.

The EVAO was the value per property on their livestock enterprise; if it was above $40,000 per year, it would be included in the survey. If it was below, it would be excluded. In 2016, the previous EVAO of $5000 per property was lifted to $40,000.

The following is a list of the new ABS data collection points:

• Historical Agricultural Census and Agricultural Commodity Survey data (to identify sub-population relationships that are relatively stable through time by State /Territory)
• ABS Trade data (for live cattle exports)
• ABS Slaughter survey on numbers of cattle and sex at slaughter. (A portion of these is reallocated back to the state of production based on livestock movement data)
• ABARES Farm Survey, which gives us rates for mortality on farm and mating (rates of adult females joined) and calving rates (rates of calves from joined females
• BOM rainfall data (to identify relationships between changes in rainfall and calf numbers for use in predicting the latest year.

The following will be added in the future as additional collection points:

• More detailed NLIS movement data to help us understand gross and net animal transfers between locations and from farms to farms, feedlots, abattoirs or any combination of these).
• Pasture condition and growth from satellite data sources such as Long Paddock
• Aggregate farm software data on herd demographics.
• Animal weight and pricing information on cattle and feed.

Why is there such a drastic change in cattle herd numbers?

The proper size of the Australian herd has been debated for some time with what has been described as a ‘contestable space’ and, therefore, the need to reassess where the baseline is because, for many, it has been too low for too long and does not reflect the proper actual size of the herd.

Using the average of the last five years of $40,000 EVAO versus the new methodology estimates, the old baseline herd size is 24.5m head, and the new baseline is 28.81m.

The ABS said the reason for raising the herd number by more than four million head was based on the herd’s inflows and outflows and the fact that the volume of the movements was far greater than what was physically possible with the old baseline. The new baseline makes far more sense in understanding these flows within the herd.

What is important to note is that this does not change any of the percentage movements of the herd size over the last five years, so if I applied my method of determining herd size, which I had at 25.5m for 2023, this equates to almost precisely 29.88m for 2023 using the new ABS baseline methodology.

This adjustment has relieved Australia’s meat processing sector, which has made sizeable investments in upgrading and building new facilities over the last five years.

This more significant herd number does not affect past and future slaughterings. Still, it comforts those in the processing sector, with shareholders and investors, who will find that the new baseline of 28.8m is more sustainable to understand and live with than the 24.5m baseline. For many processors, this perception is important as they continue to look to invest.

Bottom line – If these trends are not compromised, I can live with a new baseline. However, I am concerned about the short-term consequences of last week’s dramatic lift by 4m in the baseline, which I discuss later in this paper.

This adjustment was possibly needed for the long term, and there is never a good time to make it. The challenge is explaining this dramatic change to customers worldwide, who may not understand that nothing has changed regarding supply from Australia.

What is the risk to the market of such a dramatic increase in cattle numbers?

There are several risks to this new baseline number of 28.2 million head.

Is there a wall of meat to be shipped?

Firstly, many customers worldwide will interpret this as a large volume of additional meat to come onto the market, which is untrue. Current volumes are healthy, but these export volumes will fall with an expected rebuild next year.

Overseas buyers will be reluctant to buy if they believe larger volumes are expected going forward and will look to move to a ‘hand-to-mouth’ approach to buying. This will be challenging to beef exporters in the short term.

If the meat is not there, then it’s not there, and eventually, overseas customers will look to buy forward, but it may take some time to convince them that there is no ‘wall of meat’ to come.

International credibility of Australia’s livestock numbers and statistics

Such a dramatic lift of four million head in the baseline implies that Australia’s historic statistics have been wrong for many years, and this is a correct assumption.

So, the credibility of Australia’s livestock number estimates has been damaged for customers worldwide and Australian industry participants who use these ABS figures to make important business decisions.

How can they trust these figures in the future? From an analyst’s point of view, as long as the trends remain the same, I can live with these new numbers. However, for those making significant investment decisions, these new numbers are a bitter pill to swallow, given that an essential, trusted source such as ABS has been wrong in their baseline for an extended period.

It will take several years for many key industry participants to trust the new methodology and higher baseline figures. The trends have not changed, but for many, convincing them will take several years before the trust is reinstated.

A reassessment of Australia’s GHG reduction target

Last month, CSIRO came out with its latest assessment of Australia’s livestock sector’s greenhouse gas emissions measurement, claiming a 78 percent reduction since 2005.

This report is based on the CSIRO estimate for beef, sheep and goats of the 2021 Australian National Greenhouse Gas Inventory. This reduction was more significant than the previous year’s 65pc reduction.

So, with an additional four million head in the system, does this mean that 16pc more cattle have compromised the reduction levels? I don’t think it has, but this would seem the obvious message on the surface.

What is important to note is the new revised estimate for 2005 for herd size. How does this compare?

ABS has sent me a preliminary herd size series dating back to 2005. When comparing herd sizes, the old baseline (24.4m) of 2005 has a 3pc herd reduction compared to 2021, whereas the new methodology has a 5pc reduction in herd size. So, this new methodology helps Australia get to its GHG target quicker.

The press and the international community will look at the absolute number of four million head and say this needs to be added to the bottom line. This is incorrect; the measurement system was designed to compare livestock numbers in 2005 with today, and given that this has a new baseline, the accounting should work in our favour.

Bottom line—This additional four million cattle is messy and confusing for many people worldwide. It undermines Australia’s credibility in global markets regarding livestock data, both in terms of export markets, statistical analysis, and our standing in GHG emission targets.

If a recalibration is needed, then so be it, there is never a good time to do it. There will be short-term collateral damage regarding markets and reputation, but these concerns will disappear with time. It’s frustrating for many of us trying to explain how such an abrupt change to the herd size has happened with our international counterparts.

In reality, the outcome works in our favour (GHG reduction) in some respects but is likely to undermine some market confidence in the short term. As stated, this will return, but it will require a lot of explanation and time.

What are the concerns with this new methodology? Will it result in more or less accurate figures in the future?

A need to anchor the new methodology to a five-year census survey

There is a risk in this new methodology because the new figures each year are based on flows in and out of the herd, not on absolute numbers that a survey would provide.

So, if a dramatic event should occur that disrupts cattle numbers, the flows will not pick this up and, as a result, will see numbers ‘go out of kilter’. So, if the previous year of data is incorrect, the following years become incorrect, with the degree of inaccuracy increasing each year after the extreme event.

An excellent example of an extreme event is the Northern Queensland floods in February 2019, which saw an estimated half a million head of cattle lost. This enormous loss would impact herd size, which may not be picked up in inflows. Another example is extreme drought conditions, which might result in large on-farm losses that go unreported.

To avoid this problem, a survey should be conducted every five years, ideally during the census, meaning 120,000 agriculture operations would be surveyed. This would anchor both previous and post-measurements based on flows. To me, a survey every five years is critical to ensure the new methodology does not ‘drift’ too far away from the actual herd size.

There is precedent to my suggestion. When discussing this need with the ABS, they agreed that a benchmarking method is needed. They said they currently do a similar process on human population statistics, where similar benchmarking occurs, providing a hard number that helps ensure the in-between years are relevant and have not drifted away from the true population numbers.

Bottom line – Ideally, with any such dramatic change in statistics, you would have run the old methodology against the new methodology over five years to ensure that the new approach is accurate. Unfortunately, we do not have that luxury.

Therefore, it is crucial to conduct a survey every five years to cross-reference the new methodology, ensure that it remains true to the old process, and ensure that figures do not drift away from the correct livestock numbers—in other words, it is important to anchor the statistics to a hard number every five years.

Does this make the old data irrelevant?

Unfortunately, the new baseline will impact previous cattle herd estimates dating back to 1965 and earlier. ABS should backdate herd size figures as far back as possible. This was discussed this week, and the date backing will likely occur to 1990, though this is still being discussed.

The reason is that in analysis, we look closely at trends, and the ability to look back at herd size and slaughter levels and compare them with today is critical in understanding different cycles in the market. This is particularly relevant with unusual anomalies such as last year’s eight months of dry conditions and large cattle sales.

This historical data becomes useless and irrelevant if we do not adjust previous cattle herds to match today’s new baseline.

ABS recognised this and sent me preliminary data from 2005, with the new baseline built into the herd size. They have asked for this to remain confidential until these preliminary revised historic numbers are finalised.

Will this be more accurate in the future?

The answer to this question depends on where you sit in the market. Processors would regard these numbers as more accurate. However, as analysts, the trend is critical; to me, these trends stay the same whether they are the old or new baseline.

Producers are concerned that, in the short term, this higher new baseline will send the wrong message to our markets and keep pricing under pressure.

As long as we have a survey every five years (or an equivalent way to benchmark) that anchors the estimates in the in-between years, then it can be accurate and, more importantly, made available in a timelier manner and, if need be, every quarter.

What are the positives of this new methodology?

There are several positives:

• This new methodology is far cheaper to produce.
• It can be done quarterly if required. One concern in the past has been the different estimates from other organisations, such as ABARES and USDA, which now use a January 1 estimate with the new methodology, an industry-wide accepted number at least twice a year or more if required.
• The new methodology is hoped to remove the ongoing negativity surrounding the actual ABS figures and rebuild some trust in ABS figures.
• The difference between MLA and ABS figures has been significant; this has not been good for the industry both within and outside Australia. This new set of figures enables MLA to fall in line with ABS.
• It now removes the difference in EVAOs that has occurred in the past. This has always been a point of conjecture between different agencies, and the decision of whether to use $40,000 or $5000 EVAO and then relate previous EVAOs to whichever one you choose has always been uncertain. All that uncertainty is now removed.

Bottom line—I think if the new methodology is done correctly with a five-year census survey to act as a critical anchor on the previous year’s estimates and that ABS can reserve the right to adjust the previous year’s figures based on the survey results, then, yes, this new methodology can be equal to the old at a much-reduced cost.

Does this methodology compromise ABS independence or influence the change to any other data collection process?

ABS independence is crucial

The ABS assures me that this new methodology does not compromise its independence.

To me, ABS must remain independent. If industry and ABS decided on statistics and values together, vested interests would always try to influence the ABS numbers to their benefit.

ABS has assured me that the data will be collected and compiled independently from various organisations and that the ABS will be the only one doing the analysis and statistical estimates.

Will any other data collection processes be changed? 

The answer to this is no; in particular, slaughtering and carcase weights will remain the same. This is a very efficient, cost-effective process that has been in place for many years and is regarded by many of us as sacred ground that should never be touched.

Bottom line: ABS’s independence is crucial, as vested interests need to be kept at arm’s length. Ensuring that slaughterings and other key statistics will not be changed is also crucial.

My only desire is that slaughterings be reintroduced monthly, as it once was. Production figures can remain quarterly, but the importance of real-time accurate slaughter numbers would be beneficial.

Conclusions – What does this mean going forward?

There is no magical extra four million head of cattle in the herd; this is a book entry and an adjustment in the paperwork. Given the significant investments and expenditures in recent years in the meat processing and lotfeeding sectors, it’s an essential adjustment for many industry participants.

It is hoped that the concerns about the herd size number can be put to bed moving forward and that the yearly livestock numbers for sheep and cattle can be restored to some credibility with this new baselining.

I would be happy to support the new methodology on the proviso that a five-year census survey could be reintroduced or an equivalent ABS way to benchmark to ensure that past estimates are accurate and that estimates in the future have a strong anchor point to determine herd size based on flows in and out of the herd.

ABS has noted that there are a range of options for this benchmarking, including leveraging existing collection activity and utilising new data capture technologies.

 

ABS will need to rewrite history on both the sheep and the beef herd sizes to make them relevant, but this is not a bad thing given the constant changes to EVAOs that have occurred over the years, which has meant different interpretations of what these mean – this is now removed with an agreed number based on the new methodology. What is critical is that the trends do not change.

Second-guessing will always remain without this supportive five-year census data or an ABS benchmarking equivalent.

The new methodology and baseline data will enable MLA to align with ABS numbers. I believe the current livestock number discord between both organisations is not a good look within Australia or internationally.

The ABS’s ability to have herd and flock estimates bi-yearly or quarterly again ensures that other organisations like USDA and ABARE are working from the same numbers that MLA is, once again removing any uncertainty and restoring confidence in the new baseline process.

There will be short-term consequences of last week’s dramatic lift by four million in the baseline; as stated, this adjustment was likely needed for the long term, and there is never a good time to make it.

I believe that Australia’s cattle and sheepmeat supply will be considerably tight in 2025, 2026, and 2027. Therefore, this is as good a time as any to make this adjustment, as customers will be chasing our product, not stepping away for the next few years.

As stated, the challenge is explaining this dramatic change to customers worldwide, who may not understand that nothing has changed regarding supply from Australia.

Contrary to popular belief, the new baselining works in our favour with Australia’s GHG reduction target.

This change in the size of the herd is, for some, a necessary evil – we need to get behind these new changes, support ABS and rebuild market confidence in the data.

The result will be better forecasts and better decisions as we move forward.