TECHNICAL NOTE DATA QUALITY


RELIABILITY OF THE ESTIMATES

1 Estimates provided in this publication are subject to non-sampling and sampling error. The most common way of quantifying sampling error is to calculate the standard error for the published estimate. This is discussed in paragraphs 6 to 9 below.

2 Estimates that have an estimated relative standard error between 10% and 25% are annotated with the symbol '^'. These estimates should be used with caution as they are subject to sampling variability too high for some purposes. Estimates with an RSE between 25% and 50% are annotated with the symbol '*', indicating that the estimate should be used with caution as it is subject to sampling variability too high for most practical purposes. Estimates with an RSE greater than 50% are annotated with the symbol '**' indicating that the sampling variability causes the estimates to be considered too unreliable for general use. These annotations have only been applied to estimates from the March quarter 2009.

3 Non-sampling errors may arise as a result of errors in the reporting, recording or processing of the data and can occur even if there is a complete enumeration of the population. These errors can be introduced through inadequacies in the questionnaire, treatment of non-response, inaccurate reporting by respondents, errors in the application of survey procedures, incorrect recording of answers, and errors in data entry and processing. Inventories data for businesses with less than 20 employees are derived and could therefore be subject to error (although this error is estimated to be less than the sampling and non-sampling error resulting from directly collecting these data).

4 Estimates for the latest quarter presented in this publication are considered preliminary and revised estimates will be released with the next issue. As discussed in paragraphs 20 and 21 of the Explanatory Notes, seasonally adjusted and trend estimates are also subject to revision as more data are revised and more data becomes available.

5 It is difficult to measure the size of non-sampling errors. However, every effort is made in the design of the survey and development of survey procedures to minimise their effects.


STANDARD ERRORS

6 The estimates in this publication are based on a sample drawn from units in the surveyed population. Because the entire population is not surveyed, the published estimates are subject to sampling error. In calculating the standard error for the statistics in this publication, the ABS would prefer to produce a smoothed standard error for the major published aggregates as this approach takes account of the variability in standard error estimates for quarterly statistics. This estimated standard error would then be used as an indication of the sampling error for the current published series. Standard errors are based upon the data in the currently published quarter. Details of standard errors are available below.


LEVEL ESTIMATES

7 To illustrate, let us say that the published level estimate for company profits before income tax is $8,900m and the calculated standard error in this case is $200m. The standard error is then used to interpret the level estimate of $8,900m. For instance, the standard error of $200m indicates that:

• There are approximately two chances in three that the real value falls within the range $8,700m to $9,100m ($8,900m ± $200m).
• There are approximately nineteen chances in twenty that the real value falls within the ranges $8,500m and $9,300m ($8,900m ± $400m).
• The true value in this case is the result we would obtain if we could enumerate the total population.

8 The following table shows the standard errors for national and state quarterly level estimates based upon the data in the current quarter.

	Company gross operating profits	Company profits before income tax	Sales of goods and services	Inventories	Wages and salaries
	$m	$m	$m	$m	$m
Mining	189	193	690	179	124
Manufacturing	185	183	1 640	933	197
Electricity, gas, water and waste services	50	39	232	15	23
Construction	298	298	2 419	na	456
Wholesale trade	270	267	2 977	1 526	192
Retail trade	76	71	1 595	612	143
Accommodation and food services	80	77	646	66	180
Transport, postal and warehousing	241	212	1 049	na	172
Information media and telecommunications	41	37	254	na	44
Financial and insurance services	209	245	443	na	219
Rental, hiring and real estate services	209	180	687	na	117
Professional, scientific and technical services	282	298	1 827	na	541
Administrative and support services	113	111	875	na	222
Education and Training	na	na	na	na	158
Health Care and Social Assistance	na	na	na	na	309
Arts and recreation services	47	43	185	na	52
Other services	102	100	683	na	210
Total	710	713	5 097	1 861	1 044
New South Wales	na	na	2 758	na	518
Victoria	na	na	2 837	na	535
Queensland	na	na	1 862	na	478
South Australia	na	na	1 074	na	233
Western Australia	na	na	2 071	na	412
Tasmania	na	na	410	na	86
Northern Territory	na	na	263	na	62
Australian Capital Territory	na	na	433	na	122
Australia	710	713	5 097	1 861	1 044
na not available

MOVEMENT ESTIMATES

9 The following example illustrates how to use the standard error to interpret a movement estimate. Let us say that one quarter the published level estimate for inventories is $90,000m, and the next quarter the published level estimate is $92,000m. In this example the calculated standard error for the movement estimate is $850m. The standard error is then used to interpret the published movement estimate of +$2,000m. For instance, the standard error of $850m indicates that:

• There are approximately two chances in three that the real movement over the two quarter period falls within the range $1,150m to $2,850m ($2,000m ± $850m).
• There are approximately nineteen chances in twenty that the real movement falls within the range $300m to $3,700m ($2,000m ± $1,700m)

10 The following table shows the standard errors for national quarterly movement estimates based upon the data in the current quarter.

	Company gross operating profits	Company profits before income tax	Sales of goods and services	Inventories	Wages and salaries
	$m	$m	$m	$m	$m
Mining	164	181	391	90	73
Manufacturing	175	181	667	379	128
Electricity, gas, water and waste services	24	28	100	4	14
Construction	280	273	1 478	na	235
Wholesale trade	239	235	1 107	638	77
Retail trade	67	68	827	362	96
Accommodation and food services	54	54	213	26	81
Transport, postal and warehousing	127	140	458	na	112
Information media and telecommunications	36	29	186	na	34
Financial and insurance services	124	175	322	na	91
Rental, hiring and real estate services	114	157	284	na	63
Professional, scientific and technical services	201	252	975	na	266
Administrative and support services	108	112	511	na	123
Education and Training	na	na	na	na	92
Health Care and Social Assistance	na	na	na	na	156
Arts and recreation services	39	40	127	na	21
Other services	79	81	366	na	82
Total	537	577	2 427	800	469
na not available

ADJUSTMENTS TO ESTIMATES

11 Adjustments are included in the estimates to allow for lags in processing new businesses to the Australian Business Register. The following table shows the adjustments made to the current quarter's original estimates in current price terms:

	September Quarter 2012
	%
Company gross operating profits	0.4
Company profits before income tax	0.4
Sales of goods and services	0.4
Inventories	0.2
Wages and salaries	0.5

12 As previously discussed, the estimates presented in this publication are partial indicators used in the compilation of the quarterly national accounts. The movements in the Business Indicators estimates will not always be the same as the movements in the comparable national accounts series but they should be reasonably consistent after taking account of differences in concepts, scope and methodology described in paragraph 26 of the Explanatory Notes. If after taking account of these differences, there are concerns about data quality and coherence, the national accounts area provides feedback to the survey area. This process may result in adjustments being applied to the Business Indicators estimates prior to release in this publication. The objective use of the national accounts framework to provide data coherence across all ABS economic statistics ensures that a common understanding of recent economic developments is presented.


SALES OF GOODS AND SERVICES TIME SERIES

13 This publication includes estimates of sales of goods and services, by industry, and by state/territory, but estimates of national total sales of goods and services are not published. Total sales of goods and services is not an adequate indicator of the performance of the Australian economy as it includes duplication; for example, goods sold by retailers may also be included in goods sold by wholesalers in the same period. However this publication does include total sales of goods and services, by state/territory, as it is considered that there may be interest in this item as a measure of relative activity. This data should be used with caution given the potential for the data to include duplication across industries.


PROFITS TIME SERIES

14 Estimates of gross operating profits are compiled by deducting estimates of items that do not involve the production of goods and services from estimates of profits before income tax. These items include: depreciation, net interest paid, net foreign exchange gains/losses and unrealised gains/losses on the revaluation of assets. These items are considered out of scope of the national accounts item gross operating surplus.

15 As indicated in paragraph 6 of the Explanatory Notes, income items (other than sales of goods and services), expense items (other than labour costs) and profits are only collected for businesses employing 20 or more persons in the Quarterly Business Indicators Survey (QBIS).


WAGES TIME SERIES

16 The Introduction of the Quarterly Business Indicators Survey from March quarter 2001 included the collection of private sector wages and salaries by industry.


PRIVATISED MARKETING AUTHORITIES

17 Three significant privatised marketing authorities came into scope of the estimates in this publication from the September quarter 1999. The introduction of these units resulted in a break in series for estimates for inventories and sales of goods and services between the June and September quarters 1999 and comparison of the series over time should be undertaken with care.

18 The methodology used by the ABS has ensured that the trend series has not been distorted by the introduction of these units, although there is a trend break evident between the June and September quarters 1999. For this reason, the trend estimates of movement have not been released for the Wholesale trade inventories, Total inventories and Wholesale trade sales series in respect of the September quarter 1999.


PRIVATISATION OF TELSTRA CORPORATION

19 Telstra Corporation was effectively privatised on 20 November 2006. For the purposes of ABS statistics this change from public sector to private sector was effective from March quarter 2007. This has impacted on some data series presented in this publication, particularly the March quarter 2007 movements. The data items affected are sales of goods and services, wages and salaries, company gross operating profits and the related profits series in the Manufacturing and Information media and telecommunications industry. The introduction of Telstra has resulted in a break in series for some series in this publication between the December quarter 2006 and March quarter 2007. The movement trend estimates, in percentage terms, have therefore not been released for the March quarter 2007.


TREND BREAK

20 In the June quarter 2010 release, trend break corrections were applied to Mining and Total company gross operating profits, company profits before income tax, business gross operating profits and Mining sales. These corrections were necessary due to very large mining commodity contract price rises that occurred on 1st April 2010. Trend break corrections were also applied to Manufacturing and Western Australian Sales of goods and services due to the movement of a major manufacturing business from the private to public sector. When a trend break occurs in a time series it is important that the trend movement estimate be treated with caution, hence the suppression of the quarterly and annual movement estimates for June 2010.