Quarterly Employment Survey
Quarterly Employment Survey – Change to Business Size Indicator: October 2003
Employee count has replaced l full-time equivalent employees as the size indicator for a business with paid employees. The size indicator is used to identify businesses in the survey population.
Results for the December 1999 to l June 2003 quarters have been revised to incorporate the effects of the change.
There have been downward revisions l to estimated totals for earnings, paid hours, and employment.
Estimates of change and averages l have not been significantly revised.
The way in which Statistics New Zealand's Business Frame (BF) is maintained has been changed recently to provide more efficient and effective updating of the characteristics of constituent businesses. The BF provides the survey frame for a number of business surveys run by Statistics New Zealand, including the Quarterly Employment Survey (QES). A key element of the change is the use of tax data to update the business size indicator of each business on the BF. Previously, this indicator was updated annually using 'full-time equivalent employee' data obtained from a frame update survey.
Now it is updated monthly, using Employee Count (EC) data obtained from employer tax returns. The business size indicator is used in the sample design of the QES. The change in BF maintenance has, therefore, necessitated an update of the QES sample design, which has resulted in some changes in the statistical output of the QES. Some of the changes reflect ageing of the sample design and would have occurred anyway as a result of a design update, regardless of which measure of business size was used.
As a result, there have been revisions to originally published estimates for the December 1999 to June 2003 quarters (inclusive). The results for the June 2003 release, published on 5 August 2003, were the last to be produced based on the use of the FTE size indicator. The change has not affected the current range of outputs published from the QES.
Benefits of the change l
The change provides a more robust and timely employment indicator on the BF, and improves the accuracy of estimates of both levels and changes in the QES. The update of the sample design has also made the sample more representative of the latest population. The sample design was last updated in 1999. An important result is more timely representation of existing businesses that had started to employ labour. This is due to monthly updating of the EC size indicator on the BF and quality gains from using EC as an employment indicator.
Changes to QES totals l
The revisions to
estimated totals for all surveyed industries combined are
not statistically significant (ie within sampling error of
the originally published estimate). Revisions at the
1-character ANZSIC industry level are generally not
significant. However, downward revision to totals for the
accommodation, cafes and restaurants; cultural and
recreational services industries; and upward revision to
totals for the finance and insurance industry are
significant. Estimated filled jobs underwent downward
revisions of 10.2 percent and 12.8 percent for
accommodation, cafes and restaurants; and cultural and
recreational services respectively. Total gross weekly
earnings and paid hours in the June 2003 quarter were
revised upwards by 10.6 percent and 7.8 percent,
respectively, for the finance and insurance industry.