Subannual Economic Statistics Redevelopment Programme Retail Trade Survey Redesign

Introduction | Efficiency and Respondent Load | Design of the Current Retail Trade Survey (RTS) | Proposed Design for the RTS | Future Developments | Providing Feedback and Obtaining Information

Introduction

Statistics New Zealand is currently undertaking a programme to redevelop its series of sub-annual economic statistics. The objectives of this redevelopment include improvements to the range and quality of outputs, increased operational efficiency and a reduction in the respondent load associated with data collection.

The first phase of this programme was successfully concluded with the redesign of the Quarterly Economic Survey of Manufacturing (QMS), which was released for the September 2001 reference quarter. The second phase was the redesign of the Wholesale Trade Survey which was completed and introduced from the September 2002 reference quarter.

The current phase in the redevelopment programme is a redesign of the monthly Retail Trade Survey (RTS). This redesign, which is outlined in more detail below, is intended for implementation during the fourth calendar quarter of 2003.

Subsequent phases in the redevelopment programme will include possible extensions to the industry coverage of Statistics New Zealand's subannual economic collections. The expansion of the sub-annual industry coverage would improve the quality of the statistics used in the calculation of quarterly Gross Domestic Product.

Efficiency and Respondent Load

As part of Statistics New Zealand's compliance cost management programme, it intends to make greater use of existing administrative data in these sub-annual statistics redevelopments.

Considerable success has already been achieved in this respect, with the redesigned QMS and WTS surveys achieving reductions in sample sizes for each of these surveys of more than 25 percent. A similar approach is intended for future redesigns of Statistics New Zealand's sub-annual economic surveys.

Design of the Current Retail Trade Survey (RTS)

The current RTS is a monthly postal survey, with questionnaires going out to approximately 7,700 respondents. The variables collected from the existing RTS are:

• income from sale of goods and services (collected monthly)
• total closing stocks (collected every third month).

Actual, seasonally adjusted and trend estimates of sales are published each month, and seasonally adjusted constant price sales are produced quarterly.

The current RTS is designed to provide statistics at the 95 percent confidence interval at the following level of (sample) accuracy:

• 2 % for total sales
• 1 % for interperiod movement in total sales
• 15 % for storetype sales
• 10 % for regional council sales
• 4 % for total stocks (quarterly)

The RTS was last redesigned in 1995. Over time, sample designs become less effective in representing the current population as a whole. While the original samples are maintained to include a representative selection of new businesses, periodically, panel samples need to be refreshed to reflect changes in the composition of the population.

Geographic breakdowns of total sales are currently produced for the following geographic areas:

• Auckland Region
• Waikato Region
• Wellington Region
• Canterbury Region
• Rest of North Island
• Rest of South Island

Proposed Design for the RTS

The range of variables, and the geographic breakdowns of total sales proposed for collection and publication has remained unchanged from the existing RTS design.

The range of storetypes for which estimates are proposed to be produced has increased, as can be seen in Table 1 below. Analytical back series on the new publication basis will be produced from the time of the last redesign (ie the May 1995 month).

Table 1

The use of administrative data, in particular GST returns, is being used within the redesigned RTS. These administrative data will be used to supplement the RTS postal survey, for small to medium-sized businesses. GST data is seen as being useful in representing the activities of 'simple' retailing units (ie those which are single-unit enterprises).

In addition to the use of this data to supplement direct collection, GST data will also be used to optimise the sample design and selection processes in an effort to minimise the compliance costs imposed by the RTS collection.

The sample design will be based on GST turnover and persons engaged, aiming to meet quality criteria for the following design variables at the 95 percent confidence level:

• 2 % for total sales
• 1 % for interperiod movement in total sales
• 10 % for storetype sales
• 10 % for regional sales
• 5 % for total stocks (quarterly)

Future Developments

Work is currently underway investigating the feasibility and options for extending the industry coverage of SNZ's range of subannual economic collections. The expansion of the subannual industry coverage would improve the quality of the statistics used in the calculation of quarterly gross domestic product.

Subsequent phases in the redevelopment programme will include investigations as to the most appropriate method of collecting additional statistics (such as interest and dividend flows) to provide institutional sector estimates necessary for the calculation of a quarterly Income measure of gross domestic product QGDP(I).

Providing Feedback and Obtaining Information

Any comments should be addressed to the project leader, Nairn.MacGibbon@stats.govt.nz by email or phone on 03 964 8954.