7. Additional Cost-Benefit Considerations
7.1 Distribution
The size of the deadweight loss/excess burden is not the only consideration important in assessing the impact of a tax. The second issue relates to the distribution, or incidence of the tax: who pays? The issue of who pays is important from both an efficiency and an equity perspective. As alluded to above, a desirable property of an efficient tax is that it is paid by those who derive the most benefit (or most use) from the item/product to be taxed. From an equity perspective, this effectively avoids one section of the population bearing an unnecessarily or unfairly high proportion of the tax.
There are two key dimensions in which we are interested. First, we are interested in the distribution of the tax in terms of producers and consumers. That is, what proportion of the tax is borne by suppliers and what proportion is borne by demanders of the good to be taxed? We are also particularly interested in the within-group distribution: who, among the group that bears the majority of the burden, are more heavily affected?
In work done on the possible distributional effects of a carbon tax on petrol consumption, it was found that the costs of a petrol tax would fall almost completely on consumers. The degree to which these costs are able to be passed on to consumers is determined by the relative inelasticity of demand for petrol versus the relative elasticity of supply of such. Due to the relatively inelastic demand for petrol, consumers bear most of the cost (price rise) of the tax (Kerr, 2001).22
7.2 Direct Effects
When looking at the direct effects of the tax in terms of consumer purchases, Kerr (2001) considered any petrol tax would be regressive in nature.23 That is, households with higher income levels spend more on petrol but less as a percentage of income, than lower income households. The regressivity of the tax was still evident after attempting to correct for household size differences. An important point though, is that excluding the lowest income group suggests a much less regressive relationship over most of the income range. This is because the lowest income group is thought to be strongly represented by the self-employed- who tend to under report income (Kerr, 2001).
There are a number of issues associated with the use of household income as an indicator of welfare.24 Kerr cites Poterba (1990) as suggesting expenditure is a more relevant measure of welfare because it represents "permanent income." People who spend a lot relative to their income generally do so because they have high savings or expect higher incomes in the future (Kerr, 2001).When expenditure on petrol is modelled as a share of expenditure in each decile, the regressivity witnessed when using expenditure on petrol as a share of income is no longer present. In such modelling, it is the middle-income earners who bear the greatest cost relative to their total current and permanent income.
A similar result is obtained in work on the distributional effects of a petrol tax in the UK. It was found that the effect of a petrol tax increase on the cost of living of the poorest households is smaller than the effect on richer households, with the greatest effect being on middle-income households (Smith, 2000).
7.3 "Indirect" Effects on Consumers
These analyses consider only the direct incidence of a petrol tax- the costs to consumers of purchasing the petrol directly. Consideration should also be given to the indirect effect through increases in prices of goods that require petrol as an input. Analysis by Casler and Rafiqi (cited in Kerr, 2001) suggests that these indirect effects reduce the regressivity of fuel taxes. That is not to say that the regressivity is removed completely, but that the purchasing behaviour of relatively richer households means that they bear a greater proportion of the indirect costs of the tax that is passed through in goods markets.
While the wider economy-wide effect on industries would require general equilibrium modelling, we can draw some broad conclusions from looking at possible price effects. The Producer Price Index, compiled by Statistics New Zealand, surveys price movements in inputs and outputs across a range of factors and industries. Diesel and petrol are major inputs into a range of industrial and manufacturing processes.
A broad survey of the weights given to commodities in the index indicates that the major sector that would feel the impact of increases in prices is Transport and Storage, where diesel has a 4.8% weighting and petrol 1.7%. This means that, all else equal, a 10% increase in the price of both petrol and diesel would see an increase in costs in this sector of around 0.65%. At a more disaggregated level, diesel and petrol have weightings of 10.30% and 4.6% respectively in the Road Transport sector. This indicates a 10% increase in price would lead to an increase in the costs of road transport of about 1.49%. Road Transport accounted for 1.1% of New Zealand GDP in 2003.
Table 12: Sectoral Inputs and Industry Size| Industry sector | Weighting in PPI (% of cost) | Cost effects of 10% increase in price of fuel (%) | Industry share of GDP ( %) |
|---|
| Petrol | Diesel |
| Road transport | 4.6 | 10.3 | 1.49 | 1.1 |
|---|
| Rail and other transport | 3.0 | | 0.3 | 0.5 |
|---|
| Water transport | 9.2* | | 0.92 | 2.4 |
|---|
| Air transport | 14.2 | | 1.42 | |
|---|
| Personal and community services | 3.4 | | 0.34 | 1.5 |
|---|
| Public administration and Defence | 2.6 | | 0.26 | 3.0 |
|---|
| Retail trade | 2.0 | | 0.20 | 5.4 |
|---|
| Wood and Paper product manufacturing | 2.4 | | 0.24 | 2.0 |
|---|
| Agriculture | 2.7 | 2.5 | 0.52 | 5.4 |
|---|
Source: Statistics New Zealand, NZIER
Other sectors where petrol and diesel are important inputs are summarised in Table 12 below. This shows that the Retail Trade sector is a large part of the New Zealand economy, and has a reasonably high proportion of input costs explained by transport fuel.25 Another sector with a relatively high fuel input, and large share of GDP, is agriculture. While this is only indicative, it gives some idea of sectors that might feel some price pressure from mandatory biofuel blends and concomitant increases in fuel prices.
The relative importance of diesel in the road transport industry also suggests that mandatory biofuel use targeted at diesel (i.e. mandatory biodiesel blends) would see measures aimed at renewable fuel use being disproportionately foisted upon a single sector of the economy and not necessarily that part of the economy which produces the most fuel emissions nor that which consumes the most fossil fuels. On the face of it, this is problematic, however it is our view that the price impacts of such measures would, on balance, be felt by consumers (i.e. households) rather than firms in the road transport industry.
In summary, while we have not undertaken detailed empirical analysis, we believe the distributional effects of price increases are likely to be more heavily felt by households than producers. This is due to the relative inelasticity of demand for petrol by consumers, and the relatively elastic supply of the petroleum industry.26 In line with other studies, the impact of the "tax" would likely be slightly regressive.
Back to Top