Chapter 10
The carbon tax modelling: deficiencies
Introduction
10.1
This chapter of the report outlines the processes that have lead to the
development of the Treasury modelling that has underpinned the government's
carbon tax. The chapter also outlines and discusses the shortcomings of the
Treasury modelling.
The carbon tax modelling
10.2
On 10 July 2011, the Treasurer, the Hon. Wayne Swan MP, and the Minister
for Climate Change and Energy Efficiency, the Hon. Greg Combet AM MP, released
the Strong growth low pollution: modelling a carbon price (SGLP) report.
The report contained the assumptions that underpin the modelling of the carbon
tax.
10.3
An update to the SGLP report was released on 21 September 2011 which
revised the policy parameters of the national and sectoral economic modelling
in the SGLP report.
10.4
The update presents two additional scenarios: one that reflects the Clean
Energy Future package endorsed by the Multi‑Party Committee on
Climate Change (MPCCC), with a starting carbon price of $23/t CO2‑e
instead of the $20/t CO2‑e modelled in the SGLP report and one that also
includes additional government policy measures.
10.5
According to the Treasury, the purpose of the modelling has been to:
... inform policy design and public discussion about carbon
pricing. Treasury has modelled a range of scenarios which explore different
environmental targets and design features of a carbon pricing scheme. The
modelling provides important insights into the economic impacts of carbon
pricing at global, national, sectoral and household levels.[1]
10.6
Notably, however, the SGLP report does not include modelling results for
the important case where Australia imposes a carbon tax but the rest of the
world does not move to introduce carbon pricing.
Process issues and the development of the carbon tax
Release of the carbon tax modelling
for public scrutiny
10.7
The Treasury has not released all of its modelling or the results of the
modelling for public scrutiny. The Treasury has stated that this would amount
to 'thousands and thousands of pages of modelling'.[2]
Nor has it released the data inputs and specifications it used to modify the models
it purchased from outside sources. A failure to do so was also criticised
during the Senate Select Committee hearing into the Rudd Government's Carbon Pollution
Reduction Scheme.[3]
Peer review and scrutiny by other
experts
10.8
During the questioning of the Treasury officers about the modelling the
issue of whether the model had been peer reviewed arose:
CHAIR: ... Did Treasury conduct any public workshops
on its modelling involving other modelling experts and allowing them to
critique Treasury's approach?
Ms Quinn: Are you talking about the analysis in the
most recent or are you talking in the broad at different times?
CHAIR: The most recent.
Ms Quinn: In terms of the update that we published
this week, no we did not have any workshops around that update.[4]
10.9
While the Treasury did not conduct peer review of its most recent
analysis,
the issue of whether such scrutiny had applied to earlier modelling was also
covered:
CHAIR: What about the 2011 main modelling document
that was released a month or two earlier?
Ms Quinn: We have conducted different types of
consultation exercises on different parts of the modelling exercise. It depends
a little bit on which component you are interested in. For example, it is very
important to discuss what the technology options are in the electricity
generation sector going forward. There is a great deal of interest in those
assumptions from the electricity generation sector and very different views in
the industry. So a consultation exercise was undertaken on that component of
the modelling. There are other parts that we also went to experts and asked
them for their input.
In terms of the analysis on the MMRF model, which you have
raised this morning, we engaged the Centre of Policy Studies to provide us with
a review of the analysis that we have done to check the technical components of
the modelling were to their standard.
CHAIR: How did you determine which aspects of the
modelling you would go through a process where you would have like public
workshops or opportunity for other modelling experts to critique the Treasury
approach?
Ms Quinn: It partly depended upon the availability of
experts outside—the types of people who were interested. It is based on our
experience to the modelling in 2008, where we also undertook consultations
before and after the modelling was released. So between 2008 and 2011 there was
quite a lot of engagement between Treasury and experts on the various elements
of the modelling. So based on all of that information we also went back to
people where we thought it would be particularly useful to get input.[5]
10.10
As the Treasury is not the only organisation that undertakes modelling
exercises for the government, a point of comparison was highlighted between the
transparency surrounding the carbon tax and the work of the Productivity
Commission. The Commission regularly undertakes comprehensive modelling
activities for the government. One example is its review of the economic
transformation package, the National Reform Agenda.
10.11
The exchange below highlights the different approaches to transparency:
CHAIR: I refer you to your answers to question 15 of
the questions you took on notice on 10 August. It is in relation to the release
of modelling results. In that answer you state:
'The set of information that has
already been provided is more comprehensive than comparable reports by other
organisations'
...
How can that be true, when the Productivity Commission has
actually made the Monash modelling so files associated with its National Reform
Agenda public? Is the Productivity Commission not a comparable organisation? Or
is their modelling not comparable, even though it used the same modelling you
did?
...
Ms Quinn: It comes down to what you mean by 'open and
transparent' in the sense that we had provided at that stage, as this says, 200
or 300 pages of the modelling et cetera.
CHAIR: Sorry. I can tell you explicitly what I mean by
'open and transparent', and that is that the Productivity Commission has
actually made the Monash modelling files associated with its National Reform
Agenda work publicly available, whereas Treasury has not.
Ms Quinn: The Monash University model is available in
the public domain through Monash University.
CHAIR: But you have made adjustments to it for the
carbon tax modelling.
Ms Quinn: That information has been incorporated into
the MMRF model that is available. Models evolve through time. They change as
people evaluate things, as they add in information et cetera. We worked with
the Centre of Policy Studies. Just to make it absolutely clear, the 2008
exercise was published by the Centre of Policy Studies. Monash University
professor Philip Adams published that analysis with Treasury. It was not
Treasury using a model without the model builder being happy and content and
actively reporting that analysis. Since that time, the changes to the structure
of the model that we may well have undertaken have been incorporated into the
MMRF model and are available to other people in the Australian community.[6]
Public access to the carbon tax
modelling
10.12
At a public hearing on 10 August 2011, one month after the release of
the Clean Energy Future Package, the following question was asked of the Treasury
officers at a public hearing:
Senator BOSWELL: Will Treasury provide independent
experts access to the modelling so that they can understand all assumptions and
parameters?
Ms Quinn: As was the case previously, Treasury has
provided comprehensive documentation, including 35 pages of assumptions, as
part of the report that is on the web page. This is the most comprehensive
documentation on modelling related to carbon pricing that is available in
Australia. We have provided detail on the assumptions that are important for
determining the results. Others are free to undertake modelling with their
assumptions as they have done.[7]
10.13
The questioning continued:
Ms Quinn: In providing information to the public
domain, we have provided a comprehensive amount of information. Treasury does
not own these models, so it is not possible for us to hand over someone else's
model. These models are publicly available. They are purchased and available
from organisations within Australia. There is nothing preventing people picking
up these models and doing modelling if they have a desire to do so.
Senator BOSWELL: So, if Professor Ergas were to go
with a cheque in his hand and say, 'I want the modelling and I am prepared to
pay for it,' it would be available to him? Is that what you are saying?
Ms Quinn: He would be able to pay for the models used
by Treasury and, yes, he would be able to receive those models.
Senator BOSWELL: Comprehensive models?
Ms Quinn: Yes, he would be able to obtain them from
the providers of those models.[8]
10.14
When told that he could purchase the models used by the Treasury from
their respective sources, Professor Ergas explained to the Inquiry why this was
not as straightforward or beneficial an exercise as it seemed:
... what you can purchase,
Senator, with respect, is two models. You can purchase a model called MMRF and
a model called GTEM. But what they have done is they have taken hulls of MMRF
and GTEM and they have then specified those. They have converted them into, as
it were, a set of fully worked out equations and into that inputted very
significant quantities of data. That then yields them these two fully specified
models and then one must have means, though it is not clear what those means
are, of rendering those models consistent with each other—in other words,
synchronising the results. Without access to the actual model, including the
datasets and the specifications, essentially what you are saying is that a
person who wanted access to a Shakespeare sonnet is perfectly entitled to
himself use the alphabet and an English dictionary. You are absolutely right,
but there is a very long way and a lot of duplication of effort, and a huge
amount of reinventing the wheel, absolutely pointlessly, that would be involved
in taking that approach.[9]
10.15
Professor Ergas explained why access to the complete model and all the
data is important to any scrutiny of the Treasury's reports:
(The SGLP Report) relies on
numerous assumptions, not less the assumption of concerted global action.
In saying this, let me
emphasise that it is absolutely fair for Treasury to have made those and other
assumptions. That is in the nature of modelling. But it is also fair for there
to be a full opportunity to assess the implications of varying those assumptions.[10]
10.16
Further to the public hearing on 10 August 2011, the issue was again
raised at a public hearing on 23 September 2011. Once again, the evidence highlighted
ongoing frustration by stakeholders about a lack of access to the modelling:
Dr Gruen: Senator, if I might try and clarify, it is
not up to us to object or not object. It is not up to us. We do not lay down
the law about what other institutions can or cannot do.
Senator BOSWELL: Dr Gruen, that is helpful to know,
but when people have turned up to purchase the model from ABARES, ABARES
officials have said that, because Treasury has made modifications to the model,
any decision to make the model available is a decision for Treasury. So what I
am asking you is: you have no objection?
Dr Gruen: In respect of that statement that you have
read out, our understanding is that it is not a decision for us, so I do not
think that statement is correct.
Senator BOSWELL: All right. So you have no objection?
Dr Gruen: Sorry, I am not trying to be difficult here,
but we do not go around having objections or not having objections.
Senator BOSWELL: Well, you are, you see, because
ABARES are saying you have adjusted the model and therefore you will not let us
sell it. So you are saying ABARES are wrong?
Dr Gruen: I am saying that the evidence, as you have
read it out, does not make sense to us. We are not making those statements to
ABARES.
Senator BOSWELL: So when people go down to ABARES—
Dr Gruen: It is up to ABARES and the government.
Senator BOSWELL: and ABARES say, 'Treasury won't let
us sell'—we have it on the Hansard now—we can put it on the counter and say,
'Treasury has no objections.'
Dr Gruen: You can put it on the counter and say that
it is a decision for ABARES and the government.[11]
10.17
Based on this evidence it is the Committee's view that the Treasury
officials provided incorrect advice to the Committee. Specifically, there was a
categorical assertion that GTEM was publicly available. In fact, it was not.
Moreover, Treasury’s further replies on this issue were not helpful. What is
being sought, is the release of the models used by the Treasury, along with all
the data, specifications and assumptions that the Treasury put into those
models, so that an appropriately qualified expert could examine the modelling
of the carbon tax undertaken by the Treasury. The Treasury officials appear to
be referring to the capacity of persons to purchase a licence to use the
software for those models, without the data, assumptions and specifications by the
Treasury officials. But even on those limited terms, the reality, quite
contrary to the evidence Treasury provided, is that the GTEM model on which it
relied is not available to third parties. As a result, other experts are not in
a position to seek to replicate and appropriately test, Treasury’s modelling.
10.18
The failure by the Treasury to release the modelling for public scrutiny
was also raised by the Centre for International Economics, in relation to an
assessment of the international abatement assumptions made by the Treasury:
This very high reliance on
the purchase of international abatement is a crucial feature of the Treasury
analysis and flows through to all aspects of the results. That is, the industry
results and the price results in particular depend upon particular outcomes in
the international market for abatement.
It is natural, therefore,
to ask how sensitive the results are to changes in cost of abatement in
different countries (as well as to the changing composition of policies in
different countries) and to any restrictions in abatement trade between
countries.
Without access to the
original model, it is difficult to undertake this analysis.[12]
Release of the update carbon tax
modelling
10.19
As outlined above, the Treasury released some of its modelling on 10
July 2011. On 21 September 2011, the Treasury released the publication, Strong
Growth, Low Pollution: Modelling a Carbon Price Update (SGLP Update). The release
of the updated modelling was announced by a joint press release issued at 9:18am
by the Deputy Prime Minister and Treasurer, the Hon. Wayne Swan MP, and the
Minister for Climate Change and Energy Efficiency, the Hon. Greg Combet AM MP.[13]
10.20
The SGLP Update was released at the same time (9:18 am[14])
as the first Joint Select Committee on Australia's Clean Energy Future
Legislation public hearing. The timing of this release was raised at the Joint
Select public hearing:
CHAIR (Ms AE Burke): I declare open this public
hearing of the Joint Select Committee on Australia's Clean Energy Future
Legislation inquiry into the Clean Energy Bill 2011 and related bills. ... We
have received a written submission to this inquiry from you. As you have all
indicated you do not wish to make opening statements we will go to questions. I
will kick off the questioning as the Treasury modelling has been released this
morning.
Senator CORMANN: That is very convenient timing.
CHAIR: It is beyond my control.
Senator CORMANN: People who were suspicious would
think the government had something to hide. It is hardly open and transparent
government to release it this late.
CHAIR: Now is your opportunity.
Senator CORMANN: We do not have a copy. Where is it?
CHAIR: We are not proceeding with this inquiry under
this—
Senator CORMANN: Have you got a copy?
CHAIR: No, I have not. All I have seen—
Senator CORMANN: How are you going to ask questions
about modelling when you have not seen it?
CHAIR: I have not got a copy. That is why—
Senator CORMANN: That is ridiculous.
CHAIR: All I have seen is the press release.
Senator CORMANN: The government is treating this
committee with contempt.
CHAIR: Fine, then we will not deal with the modelling
if you do not wish to.
Senator CORMANN: It is absolutely ridiculous.
CHAIR: I was going to ask them to give us a quick
briefing on it. All I have seen is the press release from the Treasurer this
morning.
Senator CORMANN: The Treasurer is treating us with
contempt.
CHAIR: No.
Unidentified speaker: That was minutes ago, Madam
Chair.
CHAIR: It was literally minutes ago. It came through
minutes ago so I was going to see—
Senator CORMANN: That is disgraceful.
CHAIR: Right, I am not going to allow this committee
to descend into a farce at your convenience at the outset.
Senator CORMANN: It is the government that is making
it a farce.
CHAIR: Order! I have authority as chair to exclude
people from the hearing and as members know I have no harm in doing it in my
role as deputy chair in the House. So, if you want to proceed in that manner we
will not have a hearing. If you do not want to deal with the modelling we can
wait and we will recall Treasury at a later stage. I will therefore hand to
Senator Milne to kick off with relevant questions.
Mr TONY SMITH: You confirmed at the beginning of this
public hearing that the updated Treasury modelling, which was promised to be
released with the legislation, has been released minutes before the opening of
this hearing. That has just occurred now and copies are just coming into the
room now. It is quite reasonable that members of the committee regard that as
an utter contempt of the committee and, also, disrespectful to you, Madam
Chair. You are the chair of this committee and you have said just now that you—
CHAIR: The member for Casey will well know that this
is an inquiry into the bills before us today. You have the bills—they have been
there. This is additional information that goes with the bills. I will—
Senator BIRMINGHAM: Which was promised to be released
with the bills.
CHAIR: We are not in question time. We are not in the
various chambers. We are in a public hearing and we will treat it with the
respect it deserves. In that matter—
Senator BIRMINGHAM: It would be nice if the government
treated this committee with the respect it deserves.
CHAIR: Senator Birmingham, if you let me finish. In
that matter I am going to treat you with the respect you deserve and therefore
we will not deal with the modelling today. We will recall the Treasury
officials on Monday. I think that is reasonable.
Senator BIRMINGHAM: Just what questions are we meant
to ask?
CHAIR: I thought I was assisting everybody. My
apologies![15]
10.21
The Treasury did appear on Monday, 26 September 2011 at a public hearing
conducted by the Joint Committee.
Committee Comment
10.22
The Committee is concerned by the almost complete lack of transparency
about the Treasury modelling.
10.23
The modelling relies on a suite of models, and especially (for its
assessment of the economy-wide effects) on two models – the Monash
Multi-Regional Forecasting (MMRF) model and the Global Trade and Environment
(GTEM) model – along with a data set developed by the Treasury. While the MMRF
model is available commercially, the GTEM model, developed by the then
Australian Bureau of Agricultural and Resource Economics and Sciences, is not,
though Treasury categorically claimed it was in hearings of this Committee.
10.24
As a result, and given that Treasury’s data set has not been released
publicly, it is impossible for third parties to replicate, much less stress
test, Treasury’s results. Those results would therefore not be accepted for
publication in any scientific journal, and are no more than claims Treasury
makes.
10.25
This issue is made more acute by the lack of clarity on key technical
issues. Thus, it is by no means clear (and is nowhere explained in the SGLP report)
how the Treasury has inter-worked the GTEM and MMRF models. There is, with
respect to these models, a ‘double endogeniety’ problem: each model’s output is
an input to the other. There is an almost complete lack of clarity in Treasury’s
documentation about how it has addressed that problem.
10.26
Moreover, there are issues with interpreting the differences between the
models. For example, as discussed in greater detail below, the MMRF model
imposes a cost on the diffusion of induced innovations (the output of the
Marginal Abatement Cost (MAC) curves); in the GTEM model, the standard
specification does not.[16]
However, the law of one price (which states that absent trade barriers and
other trade costs, prices for identical goods will be equalised
internationally) means that it is not possible, in a conventional trade model,
for (say) the price of a scrubber for reducing emissions at cement plants to
differ as between countries. But this must happen in Treasury’s model given the
difference in assumptions about the MAC curves. This suggests the modelling
strategy may be conceptually flawed, but the Treasury has neither explained why
that is not the case nor disclosed the information needed to assess it.
10.27
Similar issues of how the models inter-work, and what decisions have
been made as to how the output of the models is combined, affect the modelling
of the price impacts of the carbon tax. In essence, the price modelling on
which the government’s compensation package relies is based on the Price
Revenue Incidence Simulation Model (PRISMOD), which is basically a calculator
of the direct and indirect impacts of assumed input price changes. These
results are likely to differ substantially from the price changes estimated in
the general equilibrium modelling, which is used to assess the effects of the
carbon tax on real incomes.
10.28
However, the Treasury nowhere discloses the price changes from the
general equilibrium modelling, nor explains its implications for real incomes
(that is, real living standards, which are the result of changes in prices and
wages), nor reconciles the various sets of results.
10.29
Treasury seems to believe this is acceptable, given that a version of
PRISMOD was used to model the GST. The Committee believes this is incorrect, for
two reasons. First, general equilibrium estimates of the impact of the GST were
available, as well as those from input-output models such as PRISMOD. Second,
and even more important, the GST was expected to increase real incomes, as tax
reform allowed the economy to become more productive. In contrast, even in
Treasury’s modelling, it is clear the carbon tax is expected to lower wages.
10.30
As a result, assessing whether adequate compensation is being provided
requires understanding both the changes in prices and incomes. There is
therefore no sensible reason why the full outputs of the modelling, in terms of
prices and incomes, have not been released.
Regulatory Impact Statement process
10.31
As part of the legislative process, legislation that is introduced into
the Parliament is required to go through a Regulatory Impact Statement (RIS)
process. The government's Best Practice Regulation Handbook (June 2010
edition) establishes the process and requirements to be followed.
10.32
As stated by the former Minister for Finance and Deregulation,
the
Hon. Lindsay Tanner MP:
Well designed regulation has a vital role to play in
overcoming some of the problems that lead to inefficient or inequitable market
outcomes. However, ‘well designed’ is an important qualifier ‑ poorly
designed regulation may not achieve its objectives, and can impose costs on
businesses and the community more broadly.[17]
10.33
The handbook is there is to provide the impetus for agencies and governments
to improve the quality of regulation and its impact on the Australian community
and economy. Given the importance of RIS process, it is unsurprising that this
matter was investigated by the committee.
10.34
At its hearing on 16 September 2011, the committee asked about the
nature of the Department of Climate Change and Energy Efficiency's compliance
with the RIS process:
Senator FIFIELD: I ask a few questions about the
regulatory impact statement on the clean energy legislation. With such
significant legislation, as a matter of course, these statements must be done.
Has the department commissioned external work to estimate the compliance costs
on businesses of the carbon tax?
Dr Kennedy: Well, as you know there was a regulatory
impact statement completed and published as part of the introduction of the
legislation. I will just hand over to Mr White, who may know a bit more detail
about that.
Senator FIFIELD: Thank you.
Mr White: Senator, in preparing the regulatory impact
statement, the department did not commission external device for the regulatory
impact statement that was prepared this year. What was done was that two quite
significant pieces of external advice were commissioned in relation to the
Carbon Pollution Reduction Scheme proposal of a couple of years ago. And
because the underlying mechanics of the schemes are similar, in terms of the
reporting requirements, permit surrender requirements, so for business systems
those reports we used as a basis for assessing their compliance costs and the resident
was prepared this year.
Senator FIFIELD: Okay, so there has not been new
analysis done.
Mr White: The department did not commission a new
external analysis.
Senator FIFIELD: Okay. So you are relying on that 2008
work, fundamentally?
Mr White: Yes, updated by the department's own
internal analysis.[18]
10.35
It is surprising that such a significant change to the economy through
this carbon tax initiative was not subject to greater scrutiny as part of the
RIS process:
Senator FIFIELD: Okay, I will just come back to the
role of the department itself in undertaking the regulatory impact statement.
The handbook also goes on to say that:
Where possible, quantify the impacts. At a minimum
your analysis should attempt to quantify all highly significant costs and
benefits to be assessed as adequate, and (the RIS) must have a degree of detail
and depth of analysis that is commensurate with the magnitude of the problem
and the size of the potential impact of the proposal.
Now, it is hard to (think) of an economic change, a policy
change, a legislative change which would have a magnitude of impact greater
than what is being proposed with the clean energy package. This is exactly the
sort of scenario that the Department of Finance and Deregulation envisages
detail and depth being gone into with a greater magnitude than in other
situations. Given that requirement, it does strike me as suboptimal, to say the
least, that it is 2008 work for a different legislative package and a different
scheme which is fundamentally being relied upon to determine the compliance costs
and the effects on business.
Dr Kennedy: As Mr White said, much of the mechanics of
this scheme are similar to the mechanics in the former CPRS. Hence we took the
judgment that we could rely on those significant early external assessments,
and the OBPR was comfortable with that position.[19]
10.36
While the RIS does not appear to have been substantially updated since
the Carbon Pollution Reduction Scheme (CPRS), the carbon tax will introduce new
compliance costs for business. As mentioned earlier, the RIS process is
intended to contribute to the development of best possible legislation:
Senator FIFIELD: As part of the process, has the
Department of Climate Change and Energy Efficiency undertaken a quantitative
cost-benefit analysis?
Dr Kennedy: Of the carbon price emissions trading
scheme?
Senator FIFIELD: Yes. And taking into account
compliance costs as part of that.
Dr Kennedy: I suppose for this policy the Treasury
modelling is the key modelling, particularly for the cost side. What this means
for variations in growth for what it might be otherwise—
Senator FIFIELD: Let us come back to compliance costs.
The Ernst & Young work in 2008 looked at compliance costs. Has that work
been done again in relation to this package?
Dr Kennedy: As Mr White said earlier, the two earlier
studies formed the basis of the compliance aspect of our assessment of this
scheme.
Senator FIFIELD: So the answer is no, then. There has
not been freshly commissioned work to look at the compliance costs for this
scheme?
Dr Kennedy: I think Mr White has answered that
question, yes.
Senator FIFIELD: The Treasury
modelling does not cover compliance costs for business, does it?
Dr Kennedy: The costs and
benefits of an emissions trading scheme? What the Treasury modelling covers at
the macroeconomic level is the cost to the economy of imposing a carbon price.
What it does not cover are the benefits that flow from mitigating climate
change. So in a sense it is not a full cost-benefit analysis—it actually only
focuses on the economic costs and the transitional adjustment from introducing
a carbon price.
Senator FIFIELD: Under the
requirements of the Office of Best Practice Regulation, when undertaking
regulatory impact statements is it not the obligation of the department of
climate change, in this case, rather than Treasury, to undertake that
cost-benefit analysis?
Dr Kennedy: We are not talking
about compliance now, we are talking about the entire scheme and all the costs
and benefits of the scheme.
Senator FIFIELD: Yes.
Dr
Kennedy: I think
it is important that the whole of government looks at the policy. Treasury is
the place that has the resources to most carefully examine and model the costs
and transitions that come with carbon pricing. The benefit side of mitigating
emissions has been looked at by the 2008 Garnaut review, not the most recent
one, and Treasury was deeply involved in that exercise. It is really the only
exercise in Australia that has tried to quantify the benefits of mitigating
climate change or avoiding dangerous climate change. Some of those benefits are
difficult to estimate but as a policy, from a modelling perspective, the extent
of the Treasury modelling and the Garnaut style modelling is probably the most
extensive and long modelling that has ever been done on any policy. They were
looking at costs and benefits over 100 years.
Senator FIFIELD: Back to compliance costs, did the
Office of Best Practice Regulation advise you that a fresh analysis of
compliance costs was not required for the RIS?
Dr Kennedy: I will ask Mr White to comment on that.
Mr White: We discussed the question of compliance
costs analysis with the Office of Best Practice Regulation and they were satisfied
with the approach we took.
Senator FIFIELD: Would the secretary of the department
have to sign off that he was confident that all requirements had been met?
Dr Kennedy: The secretary or his delegate. I think in
this case it may have been me.
Senator FIFIELD: It may have been you?
Dr Kennedy: I sign off on a lot of things, Senator. I
am pretty sure I signed off on that one.
Senator FIFIELD: Okay, it is a pretty significant
thing to sign off on. Thank you.[20]
10.37
Leaving aside the surprising point that Dr Kennedy was uncertain whether
or not he had signed off on so important a document, it is notable that in
doing so he was unaware that the Treasury modelling explicitly does not
include the compliance costs being discussed. As that document notes:[21]
The models do not capture transaction costs in
reducing emissions, such as through regulating emission trading schemes. In the
real world, implementing and monitoring emission markets has transaction costs
...[22]
Concerns about the Treasury modelling
10.38
The Treasury's modelling of the government's carbon tax, has been widely
criticised as highly optimistic and based on implausible assumptions:
The base case is meant to
reflect a plausible reality, and I do not think anybody would imagine that the
rest of the world is going to put a carbon price in place. To me, this is more
an attempt to manipulate the outcomes of the model than to try to openly and
transparently understand the effects of a carbon tax.[23]
10.39
The Committee has identified a number of concerns about the Treasury
modelling which has been publicly released:
1.
Treasury has not performed a cost-benefit analysis of the effect of
imposing a carbon tax;
2.
it has not properly modelled the macroeconomic effects of the carbon
tax;
3.
unrealistic assumptions by the Treasury that other nations will cut
their carbon emissions in line with commitments made under the Cancun,
Copenhagen and Kyoto protocols and that credible and seamless international
trading of permits will be available on anything like the scale envisaged by
Treasury;
4.
what flows from the assumption that the economy will maintain full
employment, rather than the assumption itself; and
5.
the decision not to model results at a regional level.
10.40
The fifth issue is addressed in Chapter 6 of the report.
Cost-benefit analysis
10.41
It is of great concern that Treasury has not performed a cost-benefit
analysis of the impact of imposing a carbon tax on Australia. It is inevitable
that any change of the size of the carbon tax will have effects on the economy
as a whole.
10.42
In evidence cited above, Senator Fifield questioned the Department of
Climate Change and Energy Efficiency officers on the conduct of a cost-benefit
analysis. They admitted that no such analysis was conducted, but asserted that
that responsibility rested with the Treasury. Regardless, the important thing
is that it was not done.
10.43
Other chapters of this Report address evidence provided to the Inquiry
about the knock-on effects to the economy of the government's carbon tax,
particularly in regional areas. For example, there is evidence that it will
affect jobs, not only in energy and mining, where the effects will be direct,
but also in regions where power generation or mining are the major employers.
There has been evidence that the increase in petrol prices, when it comes in,
will affect how small businesses operate, as well as prices for towns reliant
on road transport to bring in food and groceries.
Committee Comment
10.44
The Committee believes that Treasury's failure to perform the required
cost-benefit analysis on the whole economy means that its modelling does not
provide a full picture of the effect of the carbon tax.
Macroeconomic modelling
10.45
The models used by the Treasury 'do not capture transaction costs in
reducing emissions, such as through regulating emission trading schemes'.[24]
Nor has the Treasury captured the transition costs that would be imposed on the
economy by restructuring as a result of the carbon tax. The modelling 'assumes
labour and capital adjust perfectly across industries, and it does not capture
as many of the transition costs as would be experienced in the real world'. [25]
Further, the modelling treats all household assistance as a lump sum payment
for simplicity. This means it does not reflect the distortions likely to be
caused by the actual form in which that assistance is provided.
10.46
Therefore, it must be concluded that the modelling is likely to significantly
underestimate the overall costs of the scheme.
10.47
In considering the macroeconomic costs of the carbon tax, Treasury has
modelled the impact of the tax, in ‘percentage deviation from baseline terms’.
It has shown the results of the modelling for Gross National Income per person,
the capital stock, real wages and Gross Domestic Product in Charts 5.10 to
5.13.[26]
Those charts show the percentage cost of the scheme moving at a more or less
steady pace, year after year, all the way to 2050.
Committee Comment
10.48
The Committee believes that the Treasury modelling is inadequate in its
modelling of the macroeconomic costs of the carbon tax, with the risk that
these costs are significantly underestimated. The modelling it has conducted relies
on a chain of assumptions, many of them unrealistic. That said, it is true that
all economic modelling relies on assumptions — indeed, that is the difference
between a model and a life-size reproduction. However, any serious work tests
the implications of varying at least those assumptions to which the outputs are
likely to be most sensitive. In Treasury’s report, on the other hand, there is
little sensitivity testing, even compared to the work that was done for the CPRS.
10.49
While Treasury’s modelling has been presented as reflecting the economic
impact of the government’s package of measures, the Committee feels it is far
from doing so.
10.50
To begin with, as set out in Chapter 8 dealing with the budget impacts
of the government's carbon tax, Treasury’s modelling does not cover the costs
of the additional, unfunded, outlays associated with the package. As these will
require a long run increase in taxes, they would normally be costed on a basis
that includes the distortions induced by taxes, as required by the government’s
guidelines for cost-benefit appraisal. However, Treasury simply assumes the
measures are budget-neutral, when they are not.
10.51
Moreover, Treasury’s modelling does not reflect the economic costs of
the compensation package. As indicated, it assumes households receive that
compensation in the form of lump-sum payments, which (by definition) have no
effect on decisions to work or save. However, in reality, the tax changes
introduced by the government involve an increase in marginal tax rates for many
tax earners. Additionally, the tax interaction effects between the carbon tax
and the income tax for those high income earners who do not receive
compensation amount to an increase in effective marginal tax rates. As a
result, they will distort the economy, causing added economic losses.
Treasury’s modelling ignores those effects altogether.
10.52
As for Treasury’s modelling of the electricity sector, it ignores the
proposed shut down of generating facilities, and the economic consequences of
the condition that the government has said will be imposed on generators
accessing compensation payments. The effect of those conditions will be to
distort, and hence further increase the cost, of electricity supply.
10.53
Nor does Treasury model the economic costs of the subsidies to
renewables and other industries that will be provided by the Clean Energy
Finance Corporation.
10.54
Finally, Treasury’s modelling ignores the fact that the government has
locked in compensation to households in what amounts to dollar terms by
building that compensation into the tax and benefit schedules. However, future
carbon prices, and hence revenues to government, are uncertain, which means
that fiscal risk has increased. That fiscal risk has a cost to the community
but that cost if not acknowledged, much less modelled, in Treasury’s analysis.
Unrealistic assumptions about an
international carbon trading scheme
10.55
On 10 August 2011, the Treasury officials were asked about its assumption
in relation to international action on climate change. The Inquiry was told:
The analysis we have
undertaken relating to international action on climate change indicates that
countries that have made pledges at either Cancun or Copenhagen conventions
through the UNFCCC process implement policies to achieve those pledges. For
example, the United States has pledged to reduce its emissions by 17 per cent
of its 1990 levels by 2020, and that is the assumption that we have modelled in
the 550 parts per million scenario. Where countries have identified a range in
their pledges, we have taken the low-end pledges over the period to 2020. They
are the international action assumptions that are embodied in the modelling.
For the more ambitious
international action, we have assumed that countries have to achieve the
highest of their pledges between now and 2016 and then countries have to take
greater action than is currently on the table, because there is a mismatch
between the pledges that are currently on the table and the stated agreement or
aim of parties to the UNFCCC of achieving a two degrees or less warming of the
world. There is a bit of an inconsistency at the moment between those two
pledges.[27]
10.56
The main criticism of this approach is that it there is little evidence
that the international community will, in fact, live up to all its promises on
carbon reduction. It should be stressed that this does not mean no action will
be taken to reduce carbon emissions. Rather, the assumption can be seen as at
best overly optimistic, if not naive and unrealistic.
10.57
From Australia's point of view, there are significant doubts that some
of our most important trading partners and resource competitors – China, the
United States and India – will even try to meet their commitments. This point
was made by Professor Henry Ergas in evidence to the Inquiry on 10 August 2011:
Treasury has modelled a
scenario in which the rest of the world adopts such a scheme and we do too. It
also models the somewhat irrelevant case in which the rest of the world acts
and we do not. But it has not modelled, or if it has modelled has not released,
the most relevant scenario, which is the one in which we impose such a scheme
and our major resource competitors do not.[28]
10.58
Criticism of this lack of additional modelling was also made by the
Centre for International Economics:
The price pathway, the
availability of international abatement to mitigate Australian costs, and the
competiveness implications for individual Australian industries all depend on
the simulation configuration, data and parameters embedded in the international
model (GTEM) used by Treasury. The reported analysis provides one particular
scenario for global action based around one set of parameters. Unfortunately
this provides little understanding of the overall sensitivity or risks associated
with Australian policy making in the context of global action.
Further, the simulation
configuration has a number of implicit assumptions about how a world market for
abatement may ultimately work. There is limited reported information to
understand the effect of changes to these implicit assumptions.[29]
10.59
It went on to raise another concern about the way the Treasury had
modelled the international scenario for the carbon tax:
The
Productivity Commission recently noted that most countries are not implementing
carbon policies in the most cost effective way. This means that as policy is
currently emerging, it is unlikely that the true cost of abatement will be
revealed in international markets. This is similar to having distortions to the
ideal market (as simulated by Treasury).
The
effect of these distortions is similar to increasing the marginal cost of
abatement so that distortions in the lowest cost abating countries will be most
important from a risk perspective. Within a global trading framework, the costs
to Australia are as much a function of the efficiency of other country policies
as they are a function of domestic Australian policy.[30]
10.60
The cases of Canada and the United States of America can be used to
illustrate the flaws in the Treasury's assumption. Firstly, the Treasury has
conceded that neither the United States nor Canada have met their theoretical
targets under the Kyoto Protocol, even though it was adopted in 1997 and went
into force in 2005.
10.61
When pressed in relation to action by Canada to meet its promised
targets, Ms Quinn stated '(t)he Canadian government has still maintained its
commitment to achieve its pledge of similar reductions to the United States'.[31]
10.62
However, the current Canadian government was elected on a platform of
not implementing a carbon tax. It is true that it has stated it will achieve
its abatement commitments in other ways, however, it has not shown any strong
determination to do so. In fact, according to a Report released by a research
group based at the University of Ottawa on 22 September 2011, 'Canada's federal
approach to climate change has been characterised by its changing focus,
uncertainty and lack of commitment'.[32]
10.63
The United States has taken some steps to reduce its carbon emissions but
a recent decision by the Environment Protection Agency to postpone the first
round of planned greenhouse gas regulations has been described as 'the latest
step in the Obama Administration's lengthy walk back of its promised climate
policies.'[33]
10.64
Asked about the possible effect of the United States not meeting its
Cancun and Copenhagen commitments, which are greater than Australia's it must
be noted, Ms Quinn was understandably unsure as to what might result:
In terms of the economic
modelling, if the United States was not to take action to meet their Cancun
agreement of a 17 per cent reduction relative to 1990 levels, it would then
depend on what happened in other countries. Would other countries take more
action to achieve an environmental outcome that would reduce dangerous climate
change?[34]
10.65
If the assumption that countries will meet their Cancun commitments is
optimistic, then surely the hope that they will increase carbon reduction
activities to take up the shortfall created by the United States' failure to
meet its promises is patently unrealistic.
10.66
Treasury’s modelling assumes not only that countries implement their
Cancun commitments but continue on a rising abatement trajectory even after the
Cancun commitment period ends. However, the Cancun commitments are entirely
voluntary.
10.67
In modelling the consequences of these abatement assumptions, Treasury
starts from the premise that an internationally uniform carbon price will
emerge from 2016. Treasury officials were asked about how this will happen on
10 August 2011. It stated:
CHAIR: Lenore Taylor wrote in a recent article—and I think
this is similar to what you just said:
The government says it is
not assuming countries such as the US actually have an emissions trading
scheme, but rather that they would try to reach their emission reduction
targets at a cost no higher than the international price.
Do you agree with that?
Ms Quinn: Yes.
CHAIR: That is what Treasury is assuming? That is a fair
reflection of your assumption?
Ms Quinn: What we are assuming is that there are mechanisms in
countries to achieve emissions that result in an implicit or explicit carbon
price based on those economies. It does not mean it specifically has to be an
emissions trading scheme within all countries. It is the case that we are
assuming that there is a continuation of the international offset market which
exists now in order for Australia to be able to purchase permits from overseas.
So we are assuming that there is an arrangement, either through an
international framework or through bilateral trades, such that Australian
liable entities are able to purchase offsets overseas. That is not the same as
saying that all countries have to sign up to an international binding
agreement, and it would be inaccurate to make that statement.[35]
Committee Comment
10.68
As stated above, the criticism of the assumption is not that no action
will be taken, but that less than complete action is more likely – in
particular from the major economies and Australia's direct trade competitors.
10.69
The Committee, therefore, believes it is legitimate to criticise the
failure by the Treasury to model any scenario where the international community
meets something less than all of its promised emission targets as a failure on
its part.
10.70
The Treasury assumptions that countries will implement their Cancun and
that there will be a rising abatement trajectory after the end of the Cancun
commitment period are questionable. As noted above, those commitments are
voluntary. Long experience in many other areas — including international trade
and investment, as well as myriad environmental and social areas — shows that
countries frequently under-deliver or renege on voluntary commitments. As
a result, it is at the very least imprudent to assume those voluntary
commitments will indeed be delivered, and is not the way modelling would be done
for (say) defence decisions. As for the extension of those commitments, there
is no reason whatsoever to assume that will occur.
10.71
Previously Treasury denied that this meant that an international carbon
trading scheme would be in place by then. It now accepts that there will be
some mechanism by which all the advanced economies trade emissions by 2016,
will be able to trade – be it between firms, governments or both – using a
settled carbon price. However, Treasury nowhere explains what that mechanism
is, or why it is plausible that it will indeed be in place, and capable as
acting as the means by which a uniform international price emerges, by that
date. Rather, the Committee believes it is surely obvious that no such
mechanism will be in place by that time, or indeed, any time soon after that.
10.72
To then examine how the abatement targets and associated carbon prices
translate into economic activity, Treasury relies heavily on the MAC curves
mentioned above. In essence, these act as the representation in the model of
induced technical change: that is, of innovation that is stimulated by
progressively higher carbon prices. The greater the extent to which such
innovation is stimulated, the lower will be the economic cost of reducing
carbon emissions.
10.73
In itself, the assumption that higher carbon prices will stimulate
emissions-reducing innovations is plausible. But it is apparent that the reason
this mechanism works as well as it does in free market economies is that the
price increases hold out the promise of profits to innovators who can develop
products or processes that cost-effectively economise on the input whose price
is rising. International intellectual property laws, reflected in Australia’s
international obligations, ensure innovators can charge prices for those
innovations that capture for them the social value generated by their
innovations.
10.74
The Committee believes that it is at this point, the Treasury model goes
off the rails. In effect, it assumes for the GTEM modelling — which covers the
entire world other than Australia — that the induced innovations are
effectively available at no cost. This completely unrealistic assumption (which
implies the rest of the world abandons currently binding commitments to
intellectual property laws) is hardly innocent. Rather, it understates the
extent to which carbon intensive activities must shrink to achieve emissions
reductions targets and hence reduces the world income loss consequent on
emissions reduction. In turn, higher world income improves Australia’s terms of
trade and hence the income levels we can achieve under various emissions
reductions scenarios.
10.75
Unfortunately, there is no sensitivity testing in Treasury’s report of
the implications of the assumptions made with respect to these MAC curves.
However, the sensitivity tests on this assumption reported in the CPRS
modelling suggest the impacts of varying those assumptions would be very
significant. Why that is not even mentioned in Treasury’s report is a matter of
concern. Indeed, it is not even made clear in Treasury’s report that no cost
has been imputed to induced innovations in the GTEM model (see below).
10.76
Additionally, Treasury’s modelling assumes Australia can meet its
abatement targets largely by importing abatement from developing countries.
Moreover, a substantial share of those imports is projected to come from
countries in the former Soviet Union and Asia that have low estimated abatement
costs.
10.77
However, there is no reason to believe these countries will develop
credible institutions that can assure the quality of claimed abatement efforts.
This is especially important because trade in abatement differs from
conventional trade in that both the parties to this trade have an interest in
defrauding the government (who is the sole consumer of abatement). As a result,
enforcing the quality of international transactions is especially complex and
costly. But Treasury simply assumes those difficulties away, and takes it for
granted that the seemingly intractable problems associated with corruption and
low institutional competence in these countries have been solved.
10.78
Thus, there is no testing of the implications should the world not
proceed on the path to coordinated global abatement while Australia remains
locked in to the government’s carbon tax. There is, in other words, no
modelling of the implications of unilateral abatement, or of abatement by
Australia not matched by abatement by our foreign competitors.
10.79
As for the assumptions about the availability of foreign abatement, it
is in the nature of Treasury’s modelling that abatement costs will be very much
affected by the presence or absence of low abatement cost countries. In other
words, including or excluding a source of abatement with low costs has a more
than proportional impact on the overall estimated cost of achieving a given
abatement target. As a result, it is likely that Treasury’s estimate of the
costs of the government’s proposal are highly sensitive to the assumptions
about the feasibility of buying abatement from a wide range of low cost sources
overseas. However, Treasury does not provide any sensitivity testing in this
respect.
The assumption of full employment
10.80
The modelling assumes 'in the long run that there is an adjustment of
the labour market back to a structural rate of unemployment.'[36]
In other words, full employment. The Committee accepts that this is a common
assumption to make in economic modelling, but – even leaving aside the issue of
whether or not this structural rate would be unaffected by actions to forsake
much of Australia’s comparative competitive advantage in low cost coal-fired
power generation – it does not follow that the fruit of the assumption is
beyond criticism.
10.81
Furthermore, the government continues to present what is a
Treasury assumption underpinning the modelling as if it is a modelling result.
10.82
The Treasury modelling gives prominence to the claim that '(j)obs will continue to grow under
carbon pricing. By 2020, national employment is projected to increase by 1.6
million jobs, with or without a carbon price.'[37]
This additional employment will be particularly concentrated in the
services sector.[38]
Far less prominence is given to the concession elsewhere in the modelling that
substantially lower real wages (relative to baseline) would be required to
achieve such outcomes – almost 6 per cent lower by 2050 and still steadily
declining.[39]
10.83
The SGLP Report states:
The shifts in jobs between
industries caused by pricing carbon will be small compared to those caused by
ongoing technological change and income growth. They will also be small
compared to the usual churning of employment between firms and industries every
year.[40]
10.84
It goes on to say that '(t)he
impacts of carbon pricing on output and employment growth vary widely across
sectors, with some sectors growing faster, and others more slowly.'[41]
What this fails to acknowledge is that some sectors of the economy are truly
national, whereas others are concentrated in a small number of regional areas.
This is particularly so for the industries that will be most affected by the
carbon tax – the power generation and the mining industries.
10.85
The Committee's concerns about the Treasury's assumptions concerning
employment levels are:
1.
having built the assumption into the modelling, claiming it as a result
that shows full employment does not prove that this will in fact be the case;
2.
the assumption does not reflect the medium and short term effects of the
government's carbon tax policy; and
3.
the assumption operates across the entire economy, but it does not
follow that full employment will be maintained in all regions.
10.86
The actual effect of the government's carbon tax, as distinct from its
assumed effect, has been addressed in submissions to the Inquiry from New South
Wales Treasury[42]
and the Moe and District Residents Association.[43]
It has been raised in public statements[44]
and by witnesses from regional areas in their evidence to the Inquiry. [45]
Chapter 6 examines the effect of the carbon tax on regional communities in more
detail.
10.87
Treasury provides no sensitivity testing of its assumptions about the
speed at which the labour market reverts to full employment. Additionally,
Treasury implicitly but importantly assumes the equilibrium rate of
unemployment is unchanged by the carbon tax, but this assumption is plainly
incorrect - rather, the tax interaction effects discussed below (that is, the
distortions that arise from the interaction of the carbon tax and the income
tax) must mean that equilibrium rate rises. However, the Treasury provides no
estimate of that increase or of the impact of varying its assumption on labour
market outcomes.
10.88
The Committee is of the view that Treasury’s modelling of the labour
market effects of the carbon tax is unclear to the point of being misleading.
In effect, the Treasury presents the results as if unemployment was continually
at its equilibrium rate, in other words, it appears as if the labour market
continuously provides what amounts to full employment.
10.89
However, subsequent to the publication of its report, the Treasury has
accepted that in the MMRF model it uses, it takes the labour market 5 to 10
years to adjust to a shock such as the imposition of a carbon price. But there
is no sign of that adjustment, and the associated unemployment, in its
modelling results. This suggests Treasury has altered MMRF to ensure full
employment, presumably by requiring the adjustment to be completed within a
year; but it nowhere explains how it has do so, or more important, how that can
possibly be justified.
Modelling of the effect of the
carbon tax on regional areas
10.90
The downturn in manufacturing industries is already taking a toll on
some regions of the country. In the areas most reliant on power generation and
coal mining for employment the effects of the government's carbon tax will be
accentuated. The carbon tax will also affect regional communities that are
reliant on heavy vehicle transport.
10.91
The difficulties with the assumption of rapid, frictionless adjustment
to these changes are shown up most starkly by Frontier Economics' report for
New South Wales Treasury:
This
aggregate employment result [that the effects on employment are expected
to be only modestly adverse] masks the underlying structural adjustment
necessary for the economy to achieve this moderate result, which requires
employment and other resources to flow freely between sectors and/or regions.[46]
[emphasis supplied]
10.92
It is also notable that some of the Treasury modelling explicitly does
not capture all of the costs associated with such major structural adjustment.
For example, it is stated of the GTEM model used by Treasury that:
GTEM
assumes labour and capital adjust perfectly across industries, and it does not
capture as many of the transition costs as would be experienced in the real
world.[47]
10.93
What is being assumed by the Treasury is that, even though there will be
a loss of jobs in coal mining in the La Trobe Valley, for example, more jobs
will be created in other areas and over all, employment will remain the same.
This, it should be noted, is predicated on continued population growth to 35
million. The notion that people who have spent all their lives working as
miners in Traralgon will be able to obtain work in service industries on the Gold
Coast is a long bow to draw.
10.94
This is no disrespect to those workers, but merely the statement of an
economic reality that disappears under the assumption being made by the Treasury.
10.95
None of this, of course, takes into account the knock-on effects of job
losses in regions that are heavily reliant on emissions-intensive industries.
Modelling of price impacts of the
carbon tax
10.96
In 2010 the government introduced legislation to split the renewable
energy targets scheme (RET scheme) between small-scale and large-scale schemes.
On 21 June 2010, the Minister for Finance and Deregulation, Senator the
Hon. Penny Wong, told Parliament that this change was expected to increase
electricity prices by around 4.4 per cent over four years.[48]
10.97
In December 2010, Queensland's energy regulator estimated that these
changes to the RET scheme would increase power prices in that state by 2.9% in
2011. Similarly, in January 2011 the New South Wales Independent Prices and
Regulatory Tribunal (IPART) estimated that changes to the RET scheme would push
its power prices up by 6% in 2011-12.
10.98
These figures were cited to Treasury when it gave evidence to the
Inquiry on 10 August 2011. It was then asked about the accuracy and reliability
of its modelling of power prices generally:
The point here is that
Commonwealth forecasting of the impact of policy changes on the cost of
electricity is not very good, is it? I am interested in what role Treasury had
in assessing the impact of the RET on electricity prices moving forward.
Mr Raether: From memory, the report you are referring to was
commissioned by the Department of Climate Change and Energy Efficiency. As to
that particular report, you might do best to direct questions to them. I think
that was a report by MMA—a consultancy—but I do not know whether my colleagues
want to comment on the extent to which Treasury provided comments on that
exercise.
Ms Quinn: I want to provide some additional information. What Mr
Raether says is true; it was done by the Department of Climate Change and
Energy Efficiency. The number you referred to—the four per cent—was an average
across Australia. There were variations within that across different states. It
was the case that the expectation was that, as with the carbon price, there
would be a front-loading of the increase as a result of the renewable energy
target, there would be an increase in electricity costs as people adjusted and
increases in future would be less.
...
Ms Quinn: We were talking about going from a renewable energy
certificate of very low up to a large number. That initial step change can be
quite large and then incremental changes over time are quite small. So you get
a step change in the knock-on effect of that and so the percentage increase you
get in the first year is larger than subsequent percentage changes in the next
few years. I do not think you can draw your conclusion that in Queensland it is
around 2.9 per cent and therefore the four per cent number is not accurate.
CHAIR: The four per cent number is over four years.
Ms Quinn: That is right.
...
Ms Quinn: ... the 2.9 is the first year and subsequent years are
likely to be lower than that as a result of this step change implication. The
other thing to note is that the policy framework has changed since the
modelling you referred to with the separation of the renewable energy target
into different elements, and that has implications for the costings.[49]
10.99
Treasury took the question on notice. Its response did not directly
address the question of the accuracy of its earlier predictions of the effect
of the government's changes to the RET scheme on electricity prices. Nor did it
address Treasury's role in the RET forecast. It can only be concluded that it
stands by its modelling and its prediction.
10.100 The Committee is
of the view that this modelling is unreliable. The public can have no
confidence that power prices will only rise in line with Treasury predictions
as a result of the changes to the RET and the carbon tax.
Committee comment
10.101 The Committee
believes that the evidence it has received shows that there are a number of
quite significant shortcomings with the modelling conducted by Treasury:
-
it has not modelled the quite probable scenario where Australia
imposes a carbon tax and other countries, and especially Australia’s major resource
competitors, do not;
-
it has not performed a cost-benefit analysis of the effect of
imposing a carbon tax or a proper Regulatory Impact Statement;
-
it assumes that the economy will maintain full employment;
-
its estimate of the effects of changes to the Renewable Energy
Target scheme is at odds with analysis conducted for the New South Wales and
Queensland governments;
-
it has not released any modelling of the impact of a carbon tax
on specific regions of Australia; and
-
it has not allowed public scrutiny of its full models, datasets
and specifications.
10.102 Treasury’s
modelling has not been presented in a balanced and objective way, nor has
Treasury answered the questions put to it in a manner that respects the
importance of the Senate inquiry process.
10.103 First, with
respect to the SGLP report, it is at times seriously misleading, including by
not making clear the assumptions on which the analysis rests. For example:
-
it never makes it clear how Treasury has modelled the labour market
adjustment process but nonetheless emphasises that employment continues to grow
and full employment persists;
-
nowhere in the report is it made clear that some form of
generalised international carbon trading has been assumed, and never is it
explained what the form of that trading is or how it achieves a uniform world
carbon price; and
-
the report never explains that it assumes that for the purposes
of the GTEM model, induced innovation is not fully costed.
10.104 Moreover, in
answering questions about the modelling, Treasury withheld or misstated crucial
information. Thus, as noted above, Treasury maintained categorically that the
GTEM model was publicly available, when it was well known that that was not the
case. Additionally, Treasury claimed that the MAC curves were fully costed,
saying that statements by Professor Ergas, as cited above, were incorrect. It
subsequently admitted that Professor Ergas’ statement was indeed correct.
10.105 Indeed, even the
written replies Treasury provided were in important respects seriously
questionable. Thus, in replying to Professor Ergas on the MAC curves, Treasury
said that its assumption of un-costed innovation was 'conservative', as:
raising the costs
associated with the MAC curves in the GTEM model results in a lower world
carbon price path to achieve any given environmental target as global high
emission industries reduce emissions more quickly.[50]
10.106 However, as is
clear from the sensitivity testing Treasury reported in the case of CPRS, the
economic costs of abatement efforts are much higher when the GTEM MAC curves
are properly costed, as carbon intensive production shrinks much more quickly.
By not stating this, but rather presenting its approach as ‘conservative’,
Treasury was less than entirely forthright.
10.107 A similar misstatement
occurs with respect to the assumptions about the labour market. There again, in
replying to Professor Ergas, Treasury suggested that the impact of an
ever-increasing carbon price would be slight, and that this would be clear in a
forward looking model.[51]
(A model is forward looking if market participants form expectations about the
future and act on those expectations.) This is problematic, if not entirely
incorrect, for at least two reasons.
10.108 Firstly, it is
not accepted, nor is it acceptable practice, to use a backward looking model
(as Treasury has done) but then cherry pick a single area (in this case,
employment) and pretend it can be assumed to behave as if it were
forward-looking without the other model outputs also changing. Rather,
all the modelling should be placed on a forward-looking basis if those
inferences are to have any weight.
10.109 Secondly, even putting
that aside, Treasury claims that:
... after the initial introduction of carbon pricing the year
to year movements in the carbon price are expected to be relatively small,
allowing the labour market to work through most of the initial adjustment over
time.[52]
10.110 This is, at best,
unproven, as the carbon price is increasing more rapidly than the economy’s
growth rate, while the MAC curves Treasury uses appear somewhat front loaded
(so reductions in the costs of abatement occur relatively soon). To that
extent, the rising carbon price would be pushing against ever slower reductions
in the marginal cost of abatement, necessitating increasing falls in real wages
for full employment to persist.
10.111 The Committee again
recommends as per Recommendation 1 that the carbon tax be opposed by the
Parliament.
Recommendation 4
10.112 The Committee
recommends that, should the government remain committed to proceeding with its carbon
tax, before any vote the Senate should demand that:
-
the
government release all of its modelling, including the actual models, datasets
and specifications used by the Treasury, to allow third party review;
-
the
government establish an Independent Expert Panel to review its modelling
approach and framework;
-
the
Productivity Commission be asked to undertake a cost-benefit analysis of the
proposed carbon tax;
-
the
legislation should be amended to ensure that any increase in the tax or lowering
of the emissions cap be made a disallowable instrument and to ensure that
carbon permits are not private property.
Navigation: Previous Page | Contents | Next Page