Tuesday, December 06, 2016

Initial reflections on the EY APVMA cost benefit analysis

I apologise for the delay in posting. I have been slowly working my way through the Ernst and Young report on the proposed move to Armidale of the APVMA. I have now brought the first of two posts on the move up on the New England blog, Evaluating the evaluation - EY, APVMA and the move to Armidale part one.

The report is quite long, 85 pages. I needed to understand the direction they were coming from, the assumptions used. I also needed to look at the commentary and the policy background.

Much of the debate has been framed in narrow terms with people using headline numbers from the EY report for their own purposes . I don't think that it's possible to overcome this. The EY report is not especially good, but it's not a bad report either for its type. It's just limited.

Given the stated regional development policy objective behind the move, I think three questions needed to be addressed:
  • What are the costs and risks associated with the move?
  • What are the gains from a regional development perspective?
  • Do those gains outweigh the identified costs and risks?
The EY analysis concentrates on the first question. The cost-benefit analysis is narrowly defined so that it deals with the NPV of the financial costs of the shift including certain costs to industry. Here the costs of a new building plus redundancy and recruitment costs dominate, costs focused in the first few years that more than offset later property savings. The sensitivity testing suggests a range of NPV costs over twenty years ranging from $9 to $23 million depending on the combination applied.    .

The risk analysis is okay in broad terms, but really goes off the rails when it comes to calculating potential cost of lost production, a much quoted headline number. The assumption chain involved means that the final number really has no validity. The most that can be said is that it provides a worst case number of what might happen if all assumptions were met and no remedial action taken.

The way the REMPLAN input-output model is used to calculate relative impacts on the ACT as compared to the previous Armidale- Dumaresq LGA is misleading. Among other things, it's not comparing like with like.

But accepting the $23 million twenty year NPV cost for the moment, two questions remain:  
  • What are the gains from a regional development perspective?
  • Do those gains outweigh the identified costs and risks?
I will look at these questions in my next post on the New England blog. .

Postscript

I have still to finish my second post on the move, focusing on the dynamic elements. In the meantime, here is recent press coverage:

The visit to Armidale locked the PM in on the move. I can't say the reduced immediate numbers to Armidale comes as a surprise. That was built in on practical grounds. looking at the regulatory science position, I have formed the strong view that the real recruitment problem is not the shift to Armidale, that may actually help in the longer term, but the inability to get staff to go to Canberra. Neither of the two Canberra universities appear to offer any training in this area, and we already know that university graduates from the metros are reluctant to move to Canberra.



  

16 comments:

2 tanners said...

Jim

I agree with your analysis, so far. My initial assumption is that no remedial action will be taken, because it will cost and more importantly will cause delays. The operation of the AVPMA is a secondary concern in this decision. The EY analysis is flawed already (if yesterday's CT article is to be believed) because the AVPMA is now looking at operating in several locations at once for lack of any other option, whereas EY pretty much assumed it would remain together. Obviously, that was a reasonable assumption at the time.

As I may or may not have noted earlier, this is an administrative decision by a Minister and shouldn't be a Cabinet matter, even if he needed the Finance Minister to get it across the line. I'm only interested in arguing the merits of the decision and not the process. Nor am I interested in defending Canberra - it has enough resources to compensate for this loss if it wants to.

2 said...

In a Monday-go-anywhere type of mood, I wonder if Trump may be right in saying that his unpredictability is an asset in foreign policy. Generally, by following a predictable line and then being fronted by someone who breaks the rules and quite prepared to metaphorically grab an opponent by its diplomatic short-and-curlies leaves career diplomats at a loss. It will be interesting to watch, and should be judged by the outcome, not the process.

Anonymous said...

It's Tuesday tanners. Maybe time to get out of bed and assume an upright position?

Still can't see why the option to disband/do-away-with the AVPMA was not canvassed.

kvd

2 tanners said...

kvd

There was no Monday post, so I held back. Breaking the rules, living dangerously!!

On disbanding the AVPMA, someone has to approve chemicals for use. Certainly it would be an option to roll the TGA, AVPMA and IP Australia into one body, but I'm not sure many efficiencies would be gained. But EY were only asked to examine moving AVPMA to Armidale vs not moving it. I agree that long before that, various other questions should be asked.

FYI, we felt out some possibilities of at least merging TGA and AVPMA back in the nineties. It was, to put it nicely, not politically achievable.

Jim Belshaw said...

Ah, kvd. I didn't do a Monday Forum. 2t has picked that up. I promise to do so in future!

2t, your niggling made me look at all this (APVMA) in detail! It seems clear that once the decision was made, APVMA started looking at ways to make it work. My feeling expressed in the main post was that it brought change forward. One outcome may be more work flexibility.

I can see your point on President elect Trump.

Anonymous said...

On the Alphabetsoup, tanners says "someone has to approve" and I agree that is important. But what I meant was - we are studiously duplicating departments operating in the exact same sphere of responsibility as no doubt exist in NZ USA UK Canada et al. So why can't we simply adopt a policy of accepting one of their efforts? In return, maybe we could assume some sort of quid pro quo role by maintaining another of the governmental roles, and provide them with our findings?

On Trump, I have for years followed an American blog which has as one of its newer blog tags "Trump Derangement Syndrome". Locally, I think such a tag could apply to most of the reporting and analysis coming out of the Fairfax press - and I'm staring directly at my old mate P. McGeough here. He needs to be retired, or moved to the viticulture desk. Or retired.

kvd

Anonymous said...

Also - meant to say, you blow that picture up - spectacular!

kvd

Anonymous said...

Just another early morning muse:

I was reading something which referred me to Judith Curry's blog on climate science, and one commenter used the term 'alt-science' and it got me to thinking about the two most influential suf/pre-fixes floating around - "alt-" and "-gate" - and I wondered just how close we are to the invention of a topic about which we will see a discussion devolving to "alt-topic-gate" as the ultimate epithet?

tanners is of course of the alt-AVPMA push, roundly (I suspect quite rightly) denouncing the unjustified move of this body; calling out AVPMA-gate for what it is, in other words :)

kvd

2 tanners said...

Jim,

Your work flexibility comment is grasping at straws. It is entirely inconsistent with the intent of the decision which is that the AVPMA will operate as before, but 150 km or more from any state capital in an area which has access to a University. (For reasons not readily explained this does not include Townsville and other otherwise qualifying centres.) Otherwise, my postbox solution would work. And it still doesn't address why such a decision was made about such a specialised body (one which cannot operate without its workforce, which takes years to develop) without considering the risks.

Likely outcome - many will take a redundancy, will retire and be hired back as consultants at higher wages. The AVPMA will be forced to take them and pay significant penalties to the Federal Government for hiring newly redundant staff - or wait 36 weeks and get nothing done.

kvd, I hate the alt-science term. Alt-right does refer to people who think of themselves as right wing, whereas I assume alt-science are the people who dismiss science and scientific method as soon as it gets in the way of previously held beliefs or positions. They're not alt-science, they're anti-science. While driving cars and using plastics.

Jim Belshaw said...

That is a nice photo, kvd. It is one of Gordon Smith's.

2t, your comments on AVPMA seem to have gone back to the future! I will comment further after I have completed the second post.

Anonymous said...

Just reading about the winners of the Regional Australia online awards. This and last year's finalists seem a very switched on bunch - no pun intended. But was disappointed to see that regionalaustralia.org.au has its head office as:

Level 2, 53 Blackall Street, Barton, ACT 2600

Fighting a losing fight, I sometimes think.

kvd

Jim Belshaw said...

Thanks, kvd. I hadn't seen the award information. Your last point is well taken, I fear!

2 tanners said...

BTW Jim, totally agree that the 'lost production' figure is rubbish. Even where a chemical is not released, or is delayed, ag producers usually have other options. The figure of lost revenue to chemical companies is relevant and is over and above the public cost. I note in passing that farmers, the AVPMA and the chemicals industry are identified as stakeholders. So is the Armidale council. The ACT is not. Curious omission?

Jim Belshaw said...

I agree that the lost chemical company revenue figure is important. On the assumptions used, the EY report estimated that a one year approval lag represented lost sales in the range $3-10.7 million. One of the difficulties in the analysis is disentangling the effects of the move from APVMAs existing problems. http://www.canberratimes.com.au/national/public-service/train-crash-apvma-already-struggling-with-workload-20161207-gt6g24.html.

I hadn't focused on regulatory science before this. What's now becoming clear is that the location of APVMA in Armidale combined with UNE course offerings has dynamic implications that I hadn't properly identified.

Neither ANU nor UC appear to offer courses in regulatory science, perhaps because the number of potential students are too small. If I'm right, APVMA has to recruit from outside Canberra. That may be part of its problem - people don't want to go to Canberra. If UNE offers courses with local employment, then this may well draw more from the existing if smaller pool of chemistry/ag students. The net result may actually be an increase in overall supply. We have seen this before. Ag Eco is an example.

I agree with you that ACT is an obvious stakeholder.

The Joyce/Turnbull visit to Armidale has locked the move in, I think. http://www.armidaleexpress.com.au/story/4343653/apvma-move-is-a-very-logical-connection-prime-minister-malcolm-turnbull-says-in-armidale/?preview_token=FKsdG3OfzOhQO8P76t79lLG2OH7eye&time=1481156715463

Tony Windsor pointed to the irony that the pair were visiting things due to him and funded by the Gillard Government. I think that's true, but its the way things work. Without the NBN and associated initiatives, the proposed move would not be possible.

Couldn't resist a feeling of sympathy for Mr Turnbull who in some of the clips is looking very tired and finding it difficult to respond to the enthusiasm of the UNE people he was meeting!

Anonymous said...

So, a chem company sells (I dunno) Strychnine, then along comes Roundup - but there's a 1 year delay in approval. Be interested to meet any farmer who sat on his hands for a year waiting on the new product, not using the old product.

Lost sales?

kvd

ps just grabbing those names; probably not in the same useage class - but you hopefully understand what I'm saying.

Jim Belshaw said...

I do, kvd. The way that EY calculated losses was to estimate what proportion of new chemicals represented new uses as compared to substitution. They did this by taking an assumed new use range from 5% to 15%; the remaining 95-85% as substitution was ignored