View all publications

2. The Kyoto First Commitment Period

The analysis takes a snapshot of 2011/12 as being representative of the first Commitment Period (CP1) under Kyoto, while allowing enough time for the transitory effects of policy changes to have largely disappeared.

As described above, a BAU scenario is ‘shocked’ with a number of scenarios:

  • Scenario 1 – An international carbon price of NZ$25/tonne3 with the government purchasing emission units on the world market to cover New Zealand’s excess emissions. The cost of the permits is financed by higher personal income taxes. (Note that this does not necessarily mean that tax rates will be higher than they are currently, only that they are higher than in the BAU scenario)

  • Scenario 2 – A price on carbon of $25/tonne CO2 in an emissions trading scheme covering all emissions from energy and industrial processes, with free allocation of permits covering around 90% of 2005 emissions for major emitters excluding electricity generators. Emissions of methane and nitrous oxides from agriculture are exempt. Any remaining excess emissions are covered as in Scenario 1.

  • Scenario 3 – As in Scenario 2 with a higher price of $50/tonne of CO2.

Note that although these scenarios are run as ‘shocks’ relative to the BAU, it is implicitly assumed that the various policies are implemented early enough for the economy to reallocate labour and investment in response to new price signals.

Apart from GHG emissions, we do not present the results in levels. Rather they are expressed as percentage changes in real dollar amounts relative to BAU. This reflects the strength of the model being in comparative scenario analysis, rather than in forecasting levels of economic activity.

In no scenarios do the declines in private consumption and gross domestic product relative to the BAU, imply declines relative to 2007/08.

Scenario 1

Government purchases emissions units from offshore, financed by higher personal income tax. From MFE net position report, the amount involved is $228m per annum, being 9.1 Mt CO 2 at $25/tonne.

Given no deterioration in the balance of payments, the additional offshore payments by government need to be offset by an increase in the balance of trade in goods and services of $228m.

As shown in Table 1 exports rise by 0.4% and imports fall by 0.2%, as does private consumption. The main mechanism at work here is the 0.2% reduction in the real exchange rate which enables exporters to sell more quantity, albeit at lower average prices – a movement down the export demand curves. The terms of trade fall by 0.2%.

Note that the model does not simulate the absolute level of prices. It deals only with relative prices. Thus a reduction in the real exchange rate could be manifested as either a devaluation of the nominal exchange rate or as lower nominal domestic prices and wages. Either way the international purchasing power of New Zealand households falls. Measured in world prices GDP declines by 0.2% relative to BAU.

One might wonder about the low national cost – at just 0.2% of private consumption, but the size of the ‘shock’ is not particularly large. Over the period 2008-2012 the ETS will apply almost exclusively to carbon dioxide, emissions of which in 2012 are projected in the BAU scenario at 38 Mt. The future price of carbon is unknown, but at $25/tonne the value of emissions is about $950 million. New Zealand’s gross domestic product will be over $200 billion by 2012. Thus the proportion of GDP accounted for by the value of emissions is less than 0.5%. But this portion of GDP does not just disappear. Indeed, the only bits that disappear are:

  1. the resources required to pay for the emission rights that New Zealand must purchase on the international market (analogous to giving away some of our exports);
  2. the deadweight loss that is generated by the higher taxation required.

In fact (1) does not actually cause a reduction in the volume of goods and services produced by New Zealand. It is simply that more resources need to go into exporting, leaving less for private consumption. So, lower private consumption is the manifestation of (1).

Table 1: Macroeconomic Results

 

BAU

Scenario 1

Scenario 2

Scenario 3

Govt responsible for all emissions

ETS $25/tonne. Free allocation to industry

ETS $50/tonne. Free allocation to industry

Emission units required to be purchased off shore (p.a)

 

9.1Mt

6.8Mt

5.2Mt

Private Consumption

 

-0.2%

-0.2%

-0.3%

Exports

 

0.4%

0.0%

-0.1%

Imports

 

-0.2%

-0.2%

-0.4%

GDP in world prices4

 

-0.2%

0.0%

0.0%

Real wage rate

 

0.1%

-0.2%

-0.5%

Household average tax rate

 

1.4%

-1.0%

-2.4%

Real exchange rate

 

-0.2%

0.0%

0.1%

Terms of Trade

 

-0.2%

0.1%

0.1%

CO2 emissions (Gg)

37964

0.0%

-5.9%

-10.1%

Agriculture CH4 & N2O

43715

0.1%

0.0%

0.0%

Total (Gg)

81679

0.1%

-2.7%

-4.7%

Scenario 2

An emissions trading scheme with a carbon charge of $25/tonne, government purchases units for excess emissions from offshore, financed by higher personal income tax if permit revenue is insufficient. Methane and nitrous oxide are exempt. Free allocation, in most cases equal to 90% of approximate 2005 emissions, applies to the following industries:

  • Dairy processing

  • Pulp and paper

  • Industrial chemicals (fertilizer and hydrogen)

  • Non-metallic mineral products (cement and lime)

  • Basic metals (iron and steel)

  • Oil refining

Analogously, the following industries are also 90% compensated for higher electricity prices:

  • Dairy processing

  • Wood processing

  • Pulp and paper (thermo-mechanical pulping)

  • Basic metals (aluminium)

With the exclusion of agricultural non-CO2 emissions from the ETS during the first commitment period, our analysis suggests that any potential loss in asset values is negligible, although industry aggregation in the model may understate such loss. In contrast, the potential loss of international competitiveness is not negligible (under no assistance).

Hence in our modelling we treat free allocation as an output subsidy, albeit limited to the equivalent of 90% of 2005 emissions in most cases – see box.

Table 1 shows that the carbon charge reduces CO2 emissions by 5.9% or 2.3 Mt. Thus the cost of units to be bought offshore is about $170m. And with a rising marginal cost of abatement, it is cheaper to undertake some abatement domestically than purchasing units from offshore. This represents a gain from an emissions trading scheme over Scenario 1. Might the gain be higher without free allocation?

Free allocation of emission rights may be thought of as some of the proceeds of auctioned rights being recycled back to industry. Other options for recycling include lower income tax rates and subsidies for growing trees or undertaking research into carbon sequestration. We have not undertaken a full analysis of recycling options, but consider the welfare effects of recycling via free allocation against the option of recycling via lower income taxes.

Free Allocation

Free allocation has two purposes that are often confused.

  1. Fairness: In the past businesses have invested on the basis that greenhouse gas emissions were free. Subsequently imposing a price on emissions could reduce the value of such investment, perhaps leading in some cases to stranded assets. Compensation via free allocation should be via a once-only allocation of emission permits equal in value to the change in asset value. There is no economic basis for ongoing free permit allocation. Firms that close should be allowed to sell the allowances and keep the revenue as the compensation is for a lower value of assets, not for lost production.
  2. Carbon leakage: Emissions pricing may impede the international competitiveness of some industries. If this leads to lower output from, or even the closure of New Zealand plants, offshore plants would increase production and global emission would not fall. Moreover, an industry once lost to New Zealand might never be re-established, even if at some point in the future most countries impose a price on emissions. In contrast to compensation for stranded assets, in this case free allocation needs to be tied to production as it is the potential loss of output that is the problem.

From a modelling perspective these two types of compensation should be handled quite differently. Compensation for stranded assets is a financial transaction that should not affect pricing decisions, but compensation to maintain output is effectively an output subsidy and so very much a part of production and pricing decisions.

Firstly, as noted above, the most important factor determining the welfare effects of a price on carbon is the cost of any emission units that New Zealand collectively may have to purchase offshore. An ETS means that more emission reduction occurs domestically and thus fewer allowances are required to be purchased from abroad. The next most important factor is that producers and consumes face the correct set of relative prices at the margin. Free allocation need not compromise these factors.

Most recycling options then will be a second order issue in terms of the welfare effects of an ETS. Without free allocation households incur a loss in purchasing power because of the lower New Zealand dollar brought about by the increased demand for foreign exchange – to pay for the offshore emission permits.

With free allocation the exchange rate effect is smaller because exports are maintained at a higher level, as free allocation helps to preserve competitiveness. Acting against this, however, is that households must forego some of the tax reductions that would be available if all emission rights were auctioned.

Our analysis shows that these effects are largely offsetting. That is, the loss in private consumption from an ETS is not particularly sensitive to some free allocation of emission rights. However, the model does not fully consider all relevant factors:

  1. It ignores the transactions costs of free allocation.
  2. Household taxes are modelled as simple average tax rates by household income quintile. Thus the full deadweight loss from progressive income taxation is not fully captured, especially with invariant total employment.
  3. Free allocation that is too generous could provide windfall profits to overseas shareholders.

Accordingly, we would expect that over time the welfare cost of the ETS would be reduced if free allocation of emission rights is phased out, other things equal.

Note also that irrespective of the recycling mechanism, the relative welfare gain that is associated with the introduction of an ETS (Scenario 2 v Scenario 1) is likely to be under-estimated somewhat as the model does not include the effect of reductions in emissions from activities not included in the model, such as deforestation.

A domestic carbon price does not decimate the tradable sector. For a given balance of payments constraint (as occurs here) anything that impedes the international competitiveness of the economy will in the long run be offset to at least some extent by an adjustment of the real exchange rate, either in the form of lower domestic prices or a devaluation of the nominal exchange rate.

As shown in Table 2 only Oil Refining and to a lesser extent Non-metallic Mineral Products incur falls in output. (Electricity is not a traded industry.) Underlying these reductions are increases in output prices of 0.7% and 0.4% respectively. It is unlikely that such increases endanger the overall viability of these industries. Note that for Oil Refining free allocation covers its own direct emissions from the refining process, but not the emissions produced when the refined fuels are combusted in vehicle engines. Thus its lower output is a direct result of less consumer demand for liquid fuels.

Scenario 3

As in Scenario 2 with the carbon price doubled to $50/tonne. This also doubles the value of free allocations.

The results confirm one’s prior expectation that the higher the carbon price, the greater the welfare cost of meeting a given emissions obligation. Private consumption declines by 0.3% compared to 0.2% in Scenario 2. One might have expected a larger fall, but the negative effects of the higher carbon charge are cushioned by greater domestic abatement. Emissions of CO2 fall by 10.1% compared to 5.9% in Scenario 2. Hence the cost of emission permits from offshore does not double, rising from $170m to $255m.

Note that a doubling of the carbon price delivers less than a doubling (70%) of the reduction in emissions, indicating a rising marginal cost of abatement.

The free allocations to selected industries are insufficient to offset the effects of the carbon price on other exporting industries, resulting in lower exports than in the BAU. Hence the adjustment on the external account is via lower imports and a small gain on the terms of trade. Imports fall by 0.4% relative to BAU, or double the fall in Scenario 2. Most of the fall is accounted for by lower imports of consumer goods and services.

While the drop in private consumption represents an unambiguous economic loss for New Zealand households, caused primarily by the real resource cost of purchasing emission permits from offshore, the rebalancing of government income improves the allocative efficiency of the economy – enough to prevent GDP measured in world prices from falling, but not enough to prevent private consumption from falling.

Table 2: Gross Output

Gross Output

Scenario 1

Scenario 2

Scenario 3

Govt responsible for all emissions

ETS $25/tonne. Free allocation to industry

ETS $50/tonne. Free allocation to industry

Dairy processing

0.1%

0.1%

0.1%

Wood processing

0.2%

0.1%

0.2%

Pulp and paper products

0.3%

0.9%

1.9%

Oil refining and products

-0.1%

-3.7%

-6.8%

Chemicals - industrial

0.2%

0.3%

0.4%

Non-metallic mineral products

0.1%

-0.4%

-0.7%

Basic metals

0.4%

3.3%

6.3%

Electricity generation

0.0%

-2.7%

-5.1%


3 The lowest price of emission used in this report is $25/tonne. It was thought that using lower prices would not be useful as many of the metrics could end up being rounded to zero.

4 GDP in world prices is considered to be a better indicator of GDP in this case, than if specified in NZ$, because it includes the effect of changes in New Zealand’s real exchange rate.