|COMMONWEALTH GRANTS COMMISSION
PROPOSED METHODS FOR REVENUE ASSESSMENTS
STAFF DISCUSSION PAPER (CGC 2007/03-s)
New South Wales TREASURY SUBMISSION
New South Wales supports the principles of simplification, transparency and the use of broad indicators in developing revenue assessments in the 2010 Review.
We agree with the use of compensation of employees (COE) as the broad indicator of the payroll tax base. We also agree with removing government administration, health and education from COE.
The Commission has stated that it is its intention to retain a threshold in this assessment. This raises two issues:
The average (policy neutral) size of the threshold; and
How to obtain the necessary data to derive the share of a state’s payroll which is above the threshold in a manner which is reliable, simple and accurate.
The average size of the threshold is relatively straightforward and can be derived using a similar method to that currently used.
However, there are significant difficulties in deriving data on the share of a State’s payroll which is above the threshold.
The weights which the Commission is proposing to use are based on data showing the proportion of firms employing more than 20 employees and the national industry concentration.
These weights, designed to reflect concentration and employment in the various industries, and to act as a broad indicator of the effect of the payroll tax free threshold, appear arbitrary.
Similarly, a general weight of 0.9 proposed for the ACT to account for the disproportionate influence of the Commonwealth government also appears arbitrary.
We have concerns with the use of national weights because of the underlying assumption that all States have the same proportion of small firms in each industry.
In the Commission’s Technical Paper (Paragraph 21), it was noted two States have a different split of large and small firms.
In the industries with a below average concentration of large firms in the ACT, there were a higher proportion of people working in firms employing 20 people or more.
In the industries with an above average concentration of large firms in the Northern Territory, there were a lower proportion of people working in firms employing 20 or more people.
This shows the use of a single weight based on the proportion of employment in large firms may not be appropriate as it may not adequately reflect differences in the employment structure across jurisdictions. The Commission should investigate this further.
Industry structure and the relationship between industry structure and average firm size can vary over time and over the course of the economic cycle.
The assessment needs to reflect the impact of changes in the industry structure, or changes in the relative importance of different industries across jurisdictions over time. These differences can have a substantial effect on States’ capacity to raise revenue from the payroll base.
We agree with Paragraph 36 that it is worth refining the weights by examining employment and/or turnover data for both small and big firms using updated ABS Business Register data and SET or EEBTUM data at unit level or econometric means using past data as a guide before a final decision is made.
The current approach makes use of a number of surveys in determining the proportion of compensation of employees that is assessable for the determination of payroll tax. There are some concerns over the reliability of some of these data sets and the use of judgement.
The alternative being considered in this assessment is complex and is not necessarily any more reliable or transparent given the assumptions being made in an attempt to determine the wages that are assessable above the threshold.
This assessment needs to be considered further. In particular, the use of a threshold, and the manner in which an allowance is made for the wages that are paid below the threshold need to be considered further.
New South Wales remains to be convinced that the proposed method is more reliable, simple and transparent than the existing method.
We agree with Paragraph 41 that conceptually State revenue office (SRO) data for non-principal residential (NPR) land should represent a better measure than the use of a proxy base. However, we also acknowledge that there are significant difficulties in obtaining nationally consistent data.
We believe the focus should be on the re-examination of using NPR land value data in the 2010 Review, provided previous issues of data availability, comparability and consistency are satisfactorily resolved. An expert should be engaged to assist in improving the availability and comparability of the land data.
Broad indicators such as population and/or employment density are useful as an additional check.
We believe the effect of interstate differences in the land tax rates should also be considered before a final decision is made on the appropriate base for land tax.
Land tax rates affect the values of land. Higher land tax rates would reduce the returns to commercial investors and thus would reduce the prices for land that these investors are willing to pay.
The Commission should adjust land values for the effect of the differences in tax rates on land valuations.
The tax rate structure that is applied to the land tax base reflects an element of policy choice by the States.
New South Wales applies a flat rate land tax above the threshold. Some other States apply a progressive tax rate above the threshold. New South Wales also has one of the highest thresholds. The application of adjustments for both the threshold and rate progressivity doubly disadvantages New South Wales.
Moreover, the application of a value distribution adjustment requires interstate data to be comparable for each value range. This, however, requires addressing problems relating to aggregation, strata title as well as differences in valuation dates.
New South Wales considers that value distribution adjustment is a selective aspect of States’ detailed tax design and is inconsistent with the terms of reference to simplify the assessment structure. We, therefore, support the removal of value distribution adjustment from the 2010 assessment framework.
New South Wales supports the Commission in engaging an expert to improve the comparability of land value data across Australia.
New South Wales broadly supports investigating the use of more objective, broader measures for this revenue base. However, investigation into other broader measures to date indicates value of transactions data remain the better measure of this revenue base.
Value of transactions data are a better measure of this revenue base because:
The data measure the base directly and comprehensively;
Differences in the way States report the data can be adjusted to ensure better comparability across jurisdictions.
We are concerned with the proposed use of secured housing finance loan commitments (SFC) as a broad indicator of the transfer duty base because:
The series does not capture commercial activity;
The series does not cover housing transactions that are not financed by a loan (for example, equity financed loans);
The series is less volatile than the actual transfer duty series;
The series may not capture large one-off transactions;
The series does not reflect differences in States’ residential and commercial property cycles and the relative sizes of swings in the revenue base.
Underlying the use of the SFC series is the assumption that the value of residential transactions and other transactions are well correlated. We do not consider this to be the case. The Commission needs to investigate further whether this assumption holds across jurisdictions.
The Commission proposes to introduce weights to the SFC to capture borrowing behaviour, extent of investor housing and turnover of properties. However, these weights are derived by judgement, albeit based on three measures of indicators of borrowing behaviour. The use of judgement should be avoided wherever possible. Any assessment should be based on a sound method and reliable data, consistent with the terms of reference for the 2010 Review.
We remain concerned with the stability of the weights. Further analysis is needed to verify their stability over time.
For the reasons discussed above, we consider value transactions data are better and should continue to be used.
Total insurance premiums appear to be a reasonable indicator. Currently, five States impose a levy (either a flat fee or as a percentage) on CTP insurance. The duty on CTP insurance is generally lower than the stamp duty on general insurance.
SURCHARGE/LEVY ON MOTOR VEHICLE THIRD PARTY VEHICLE INSURANCE
10% stamp duty on insurance premium charge.
(That is, private motor vehicle (high risk), $347.00 premium, insurance duty $34.70.)
A $7 insurers surcharge applies on CTP policies.
10% stamp duty on insurance premium.
9 monthly policy:
6 monthly policy:
3 month policy:
$6 per policy.
Source: Interstate Comparison of Taxes 20006-07 (TRP07-1)
A discount can be applied to the CTP insurance premium to reflect their lower effective tax rate.
Similarly, since most States do not impose stamp duty on workers compensation insurance, this component should also be excluded from the insurance base.
In assessing States’ capacity to raise revenue from mining royalties, the Commission aims to measure a revenue base that reflects the value and volume of minerals production that are subject to mining royalties, assuming the average policy is applied in all States.2-03 2003-
New South Wales does not support the value of production approach as it makes no adjustments for relative differences in the cost structures of States.
It is the overall profitability of the industry that is the best indicator of the capacity of States to generate mining revenue. States need to consider the industry’s capacity to absorb taxes when they set their revenue policies as the mining sector is highly reliant on exports.
Differences in extraction costs also reduce companies’ capacity to earn profits and hence lessen a State’s capacity to raise mining royalties.
Number of Coal Mines in NSW, by Mine Type − 1999-00 to 2005-06
For example, the NSW coal mining industry has a larger share of underground mines, which are more difficult and more costly to operate.
New South Wales, therefore, supports an assessment based on profitability as it recognises differences in the costs of production.
Disaggregating the assessment into an energy and non-energy minerals industry group would not necessarily address the cost differences. This is because industries are not uniformly profitable across all jurisdictions.
Out of the two proposed value of production measures, we prefer the one with the data sourced from ABARE. ABS data are more from an industry view, which can include activities that are not mining activities so long as the principle activity of the organisation is mining. ABARE data are more resource focused and so are closer to the base on which the royalties are imposed.
New South Wales does not oppose the inclusion of SPPs in lieu of royalties for uranium and offshore oil and gas as separate assessments, on an actual per capita basis.
5-06 (p) 2000-01 2001-02 2002-03 2003-04 2004-05 2005-06 (p)
Motor Vehicle Taxes
The use of new vehicle sales and stock of vehicles on register in each State appear to be reasonable broad indicators of this revenue base.
New South Wales does not object to combining the three indicators using weights that reflect the contribution of each vehicle type to total revenue, provided the weights are appropriate and include some form of updating mechanism.
Neither the Discussion Paper nor the Technical Paper indicates what form any updating mechanism would take.
It is unclear why the Commission proposes to add revenue from buses to the heavy and articulated trucks categories (Technical Paper Paragraph 14), when in Technical Paper Paragraph 9, the Commission stated that “unpublished data from the Motor Vehicle Census indicated that over half of the vehicles included in the buses category...are light vehicles”. The regrouping of these vehicles should be applied on a consistent basis.
Similarly, adding revenue from trailers to articulated trucks on the basis of a high correlation between trailer numbers and articulated trucks is rather arbitrary. Where the vehicles, whether it be a trailer or a special vehicle need to be categorised into other vehicle categories, they should be broadly determined by their weight characteristics and based on materiality grounds.
Long Combination Trucks
In the case of long combination trucks, the Commission has indicated it prefers to increase the weight for heavy rigid trucks and articulated trucks, each by 3 multiplicatively. This is based on judgement. We consider judgement should be avoided wherever possible.
Introducing a separate assessment of long combination trucks, while more complicated, is more transparent than the application of an arbitrary weight, and would be based on a sounder method, provided it satisfies the materiality threshold.
For the reasons discussed above, we do not support the proposed effective weights as contained in Paragraph 20 without some further analysis and explanation.
New South Wales agrees that an adjustment for the number of Commonwealth, diplomatic and consulate vehicles is not warranted on the grounds of materiality and simplicity.
New South Wales supports an equal per capita assessment for this category.
The equal per capita method is the most appropriate basis for other revenues and contributions from trading enterprises where State policies do have a substantial influence on the capacity of States to raise revenues from these sources.
Gambling policy differences between the States mean that gambling expenditure is not a sufficiently policy neutral measure of the revenue base.
Some States suggest the use of HDI as an alternative measure for gambling revenue. We consider that the best policy neutral measure for this revenue base is an equal per capita assessment.
People with high disposable incomes do not necessarily have a higher propensity to gamble. Also, it is not clear whether these people would choose to spend the same proportion of their disposal income on gambling as people in other income brackets.
The Commission’s own analysis has also indicated that gambling tax assessment would not meet the new assessment guidelines, and therefore, gambling should be treated equal per capita.
Fire Services Levy
In New South Wales, fire services levy is raised from insurance companies, local councils and State Budget, with the bulk of revenues collected from the insurance industry.
Similarly, Victoria raises its revenue from the insurance industry, local council and from its own Budget. Other States impose levies based on property and location.
Given the different means of raising this revenue source, we consider it is simpler and more consistent to assess it in the miscellaneous revenue category, on an equal per capita basis.