Wholesale Trade Survey: September 2012 quarter

Period-specific information
This section contains information about data that has changed since the last release.

• Measurement errors
• Postal response rate  

General information
This section contains information about data that does not change between releases.

• Population
• Industry descriptions
• Sample design 
• Sample maintenance 
• Sample reselection 
• Measurement errors 
• Seasonally adjusted series 
• Estimated trend 
• More information

Period-specific information

Measurement errors

All statistical estimates are subject to measurement errors. These include both sample errors and non-sample errors. In addition, we apply imputation methodologies to cope with small firms and non-response. These measurement errors should be considered when analysing the results from the survey.

Sample errors

At the industry level, the following sample errors were recorded in the September 2012 quarter, at the 95 percent confidence interval limit.

The postal survey is designed to give statistics at the following levels of accuracy (at the 95 percent confidence interval):

• 5 percent for operating income and stocks at the total wholesale trade level
• 10 percent for operating income and stocks at the published industry level.

This means, for example, that there is a 95 percent chance that the true value of total wholesale trade operating income lies within 5 percent of the published estimate.

Imputation

Small firms
Small- to medium-sized firms are generally not surveyed. Instead, we model their variables from administrative data (GST and employer monthly schedule) sourced from Inland Revenue. Ratios calculated from the postal sample units are applied to the administrative data to provide an estimate of their variables.

Non-response imputation
Although we make every attempt to achieve a 100 percent response rate, in practice this does not occur. We estimate values for non-responding businesses by a number of methods, including:

• regression imputation
• historic imputation
• mean imputation.

Regression imputation involves estimating the variable of interest from the unit's administrative data (GST sales), based on the relationship shown by similar businesses. Historic imputation involves multiplying their response in the previous period by a non-response factor. The non-response factor is the average movement over the quarter by similar businesses. Mean imputation involves estimating a value for a unit by using the average value for a set of similar businesses.

Postal response rate

The response rate achieved in the September 2012 quarter was 91.4 percent. The Wholesale Trade Survey has a target response rate of 85 percent. The response rate describes the proportion of operating income that was provided by actual survey responses. Note that we calculate this response rate from the postal sample data only.

General information

Population

The target population for this survey is all kind-of-activity units (KAUs) operating in New Zealand that are classified as Wholesale Trade (Australian and New Zealand Standard Industrial Classification – ANZSIC06 Division F) on Statistics NZ's Business Frame.

Industry descriptions

A KAU is included in an industry based on its main activity for its operating income.

The six industries are defined as follows:

Sample design

We stratify the survey population according to:

• industries defined by the ANZSIC-based NZSIOC classification
• size (in terms of the rolling-mean employment number)
• turnover (annualised GST sales).

Each NZSIOC industry classification contains between two and four substrata. Because of the contribution large units make to the economic activity within each industry group, we include them all in the sample. A portion of the remaining medium to large units is also included in the sample. In addition, we model small- to medium-sized businesses' data from administrative data (GST and employer monthly schedule) sourced from Inland Revenue.

All wholesaling KAUs belonging to a selected 'enterprise' are included. We select about 650 units from the entire population for the postal sample, and the data for approximately 35,000 units is modelled from tax data. 

Sample maintenance

Sample maintenance is the process we use to maintain the sample over time, to reflect births, deaths, and other structural changes identified on the Business Frame. The information for Business Frame changes can be from a variety of sources, including GST registrations and respondent contact.

We identify new enterprises when they register for GST. Once a quarter, the new enterprises are selected into the sample using the same criteria as for the original sample. These are referred to as 'births'. When an enterprise ceases trading, its wholesaling KAUs are removed from the survey. These are referred to as 'deaths'.

Enterprises can also enter or leave the survey sample if they are reclassifications from another industry to wholesaling. Reclassifications occur when an enterprise changes its main form of activity (eg from manufacturing to wholesale trade). These are usually identified in the Annual Frame Update Survey conducted in February each year.

Sample reselection

We reselect the sample for the Wholesale Trade Survey each quarter to ensure that the sample reflects changes occurring in the wholesale trade population.

Measurement errors

Errors in the survey are divided into two classes: non-sampling error and sampling error.

Non-sampling error

Non-sampling error includes errors arising from biases in the patterns of response and non-response, inaccuracies in reporting by respondents, and errors in recording and coding data. The size of these errors is difficult to quantify. We revise data if significant errors are detected in subsequent quarters.

Sampling error

Sampling error is a measure of the variability that occurs by chance because a sample, rather than an entire population, is surveyed.

Seasonally adjusted series

The X-12-ARIMA package is used to produce the seasonally adjusted estimates and trend estimates for sales in all subdivisions. Seasonal adjustment aims to eliminate the impact of regular seasonal events (such as annual cycles in agricultural production, winter, or annual holidays) on time series. This makes the data for adjacent quarters more comparable.

We revise all seasonally adjusted figures each quarter. This enables the seasonal component to be better estimated and removed from the series.

Restructuring within the dairy industry affected the grocery, liquor, and tobacco products industry series in the September 2002 quarter. In order to maintain the long-term continuity of the seasonally adjusted and trend series for primary product, food, and total wholesaling, we adjust the actual series before the seasonal adjustment program is run. This adjustment to the actual series removes the discontinuity in the series.

The X-12-ARIMA seasonal adjustment package is very robust. However, problems occur when there is an abrupt change in the seasonal variation, as with other seasonal adjustment packages.

Estimated trend

For any series, the survey estimates can be broken down into three components: trend, seasonal, and irregular. While seasonally adjusted series have the seasonal component removed, trend series have both the seasonal and the irregular components removed. Trend estimates reveal the underlying direction of movement in a series, and are likely to indicate turning points more accurately than seasonally adjusted estimates.

We calculate the trend series using the X-12-ARIMA seasonal adjustment package. They are based on a five-term moving average of the seasonally adjusted series, with an adjustment for outlying values.

Trend estimates towards the end of the series incorporate new data as it becomes available and can, therefore, change as more observations are added to the series. Revisions can be particularly large if an observation is treated as an outlier in one quarter, but is found to be part of the underlying trend as further observations are added to the series. Typically, only the estimates for the most recent quarter will be substantially revised.

More information

See Wholesale Trade Survey for more information. 

Liability

While all care and diligence has been used in processing, analysing, and extracting data and information in this publication, Statistics NZ gives no warranty it is error-free and will not be liable for any loss or damage suffered by the use directly, or indirectly, of the information in this publication.

Timing

Our information releases are delivered electronically by third parties. Delivery may be delayed by circumstances outside our control. Statistics NZ accepts no responsibility for any such delay..

Crown copyright©

This work is licensed under the Creative Commons Attribution 3.0 New Zealand licence. You are free to copy, distribute, and adapt the work, as long as you attribute the work to Statistics NZ and abide by the other licence terms. Please note you may not use any departmental or governmental emblem, logo, or coat of arms in any way that infringes any provision of the Flags, Emblems, and Names Protection Act 1981. Use the wording 'Statistics New Zealand' in your attribution, not the Statistics NZ logo.