View all publications

1. Methodology

We use a general equilibrium (GE) model to analyse various scenarios on how New Zealand might fulfil its international emission reduction obligations. A general equilibrium (GE) model is a set of equations that describe how the economy behaves, but some aspects of the economy are especially difficult to simulate. Examples include the government and the total level of investment. Hence in order to close the model – render it solvable – some extra assumptions are required.

A ‘business as usual’ (BAU) scenario is developed which represents a picture of the economy and emissions without any carbon charges or international emission obligations. The BAU is not necessarily the most likely forecast of what the economy might look like. Rather it is intended to be a plausible projection of the economy that can constitute a frame of reference against which other scenarios may be compared.1 The BAU does not take into account any of the possible climate change-related costs associated with adopting this scenario, such as trade barriers that might arise from non-participation in global efforts against climate change.

The model is then ‘shocked’ with a number of scenarios, described in the following section. In all scenarios the following are held constant at whatever emerges under BAU:

  1. Total employment, wage rates endogenous.
  2. Total capital stock, user costs of capital endogenous.
  3. Balance of payments as proportion of GDP, real exchange rate endogenous.
  4. Fiscal surplus, personal income tax rates endogenous.

The first two macroeconomic closure rules imply that the overall level of resource use in the economy is not dependent on climate change policy. Other closure rules are possible. For example instead of fixed employment, wage rates could be fixed at BAU levels. This implies, however, that the long run level of total employment is driven more by the price of carbon and energy than by the forces of labour supply and demand – an unlikely state of affairs.

The third rule ensures that the costs of meeting New Zealand’s emission obligations are not met simply by borrowing more offshore, as this is not sustainable. Relaxing this constraint would mean that in the long term New Zealand could run a larger external deficit than it other wise would – not a view likely to be shared by foreign lenders and investors.

The fourth rule prevents the results from being confounded by issues around the optimal size of government. An increase in government revenue from a carbon tax or auctioned emission permits is not a reason for enlarging government as proportion of GDP. However, other closure rules such as revenue recycling via lower corporate taxes or debt repayment would also meet this objective. Raising spending on say health, would not. If it is believed that government should be larger, then this scenario should be investigated in its own right; it is unlikely that a carbon charge is the most efficient way of doing this.

Taken together the closure rules as specified above enable us to analyse the effects of emissions mitigation measures on allocative efficiency, the terms of trade, the real exchange rate and so on, and through these variables the effect on welfare measures such as private consumption. If one wishes to claim that say, the total level of employment is indeed determined by the carbon price, then such a scenario can easily be examined as a sensitivity test by changing the labour market closure rule.

A Role for Judgement

None of the carbon price scenarios discussed in this report represent particularly large shocks. Even at $100/tonne, the cost of offshore emission units is only a few percent of GDP – not too different from the size of the recent increase in oil prices. Arguably though, for a very large shock or for a number of simultaneous and opaque shocks, some of the above closure assumptions could become unrealistic. To some extent this is a matter of timing – over a long enough period the economy will adjust to most shocks, but to dismiss prolonged periods of unemployment or low investment as merely adjustment costs is disingenuous and disdainful.

Thus the setting of macroeconomic closure rules involves a degree of judgement about what is appropriate in a given set of circumstances. That may be seen a methodological weakness of GE modelling, but it can also be a strength as models enable one to test the importance of closure (and other) assumptions.

Infometrics (2008)2 illustrates this feature. That research sought to estimate the economic effects of a very high domestic (shadow) price of carbon as a proxy for a range of non-price regulatory interventions outside the ETS, such as the ban on new thermal base load electricity generation and the biofulels mandate.

One scenario adopted the standard set of macroeconomic closure assumptions as delineated above, while another scenario argued that the type and mix of abatement policies would lead to a lengthy period of uncertainty, lobbying, distorted investment, and slower economic growth. Neither scenario was a prediction. What they did was demonstrate that changing certain closure assumptions, notably those that are questionable under a policy mix of inconsistent carbon pricing, had dramatic results. Nevertheless it is still partly a judgement call as to whether that particular policy mix is consistent with that particular mix of closure assumptions. The value of the modelling is in assessing risks and pointing to where more debate and research is required.

The following model limitations should be noted:

  • Aggregation bias – All industries in the model represent aggregations of companies, products and processes, but even with 49 industries, aggregation bias remains. For example we cannot distinguish between the production of fertilizer and hydrogen in the Chemicals industry.

  • Lumpiness in production – The model assumes that small increments and decrements in production are possible. For industries that are dominated by a single plant dependent on economies of scale this could be unrealistic, especially with respect to increments in output. However, under a carbon charge increases in output from such industries are unlikely.

  • Pricing – Being an ‘equilibrium’ model, unless specifically altered, industries must price their output at the average cost of production. There are no long run economies of scale so marginal costs equal average costs.

  • Costs of Resource Re-Allocation – The model is an “equilibrium” model. It looks at the situation after resources have been reallocated in response to changes in relative prices and changes in policy. It does not measure transition costs. Hence short term costs to the economy may be under-stated, although by a relatively small amount in a macro-economic sense, if the economy is close to capacity.

With regard to Forestry, all model runs for 2012 are on a like-for-like basis. That is, government is assumed to hold credits and liabilities for both post 1989 and pre 1990 forests, so valid comparisons can be drawn between the scenarios. In particular:

  • Post 1989 Forests – No estimate has been made on the macro-economic effect of devolving sink credits and liabilities in this modelling.  Devolving sink credits, to the degree this will occur (as it is voluntary), represents a wealth transfer within the economy and would reduce the revenue that the model has available for tax recycling.  Importantly, the number of units that need to be purchased offshore by New Zealand, over time, would not change.  However, to the extent that liabilities on harvest of forests are reduced as a result of devolution of credits and liabilities, the macroeconomic impact of the decision to devolve sink credits and liabilities will be positive. 
  • Pre 1990 Forests – The act of devolving deforestation liabilities could see significant emission reductions over the first commitment period which would reduce the need to purchase emission units offshore.  These emission reductions have not been taken into account in this modelling. However, they would work to further reduce the macroeconomic impacts under the ETS.

1 The model’s projections of the economy in 2011/12 or 2024/25 are similar to those produced by the MED in New Zealand’s Energy Outlook to 2030. Without a price on carbon, emissions are about 10% above MED/MfE gross projections, (due primarily to faster growth in emissions from agriculture and transport), but below MfEs net emissions. Regardless, given the uncertainties in these projections, the model’s emission projection represents one of the many plausible projections that could be used.

2 Infometrics (2008), Carbon Mitigation Scenarios, report to New Zealand Business Roundtable and the Petroleum Exploration and Production Association of New Zealand.