Why do the MRTEs get revised?

February 20, 2019

This article was produced by the Ministry of Business, Innovation and Employment (MBIE). The content you find below is exclusive to the Tourism Evidence and Insights Centre.

We use a modelling approach in the MRTEs which is why they are estimates

The MRTEs are our best estimations of spending in the regions, at a given point in time. Although the way we produce the MRTEs is considered world-leading, the data source is incomplete. Thus we use modelling to fill in the gaps, which is why they are referred to as estimates.

We use 3 different datasets to estimate all tourism spending:

We use these datasets to estimate the missing data (such as cash, card transactions not present in the raw dataset or TSAs that have not been released yet). These datasets all have a degree of error (uncertainty) associated with them, and contribute to the error in the MRTEs.

The MRTEs are subject to smaller revisions each month to improve the accuracy

To improve the accuracy of the MRTEs, the complete series (back to April 2008) is revised each month. This allows the most up to date information about businesses (for the ECTs), and any data from the IVS and TSA to be incorporated. Revisions are done to many statistical releases to reflect more complete information. Because of this, the MRTEs should be treated as indicative and not definitive.

The MRTEs are revised annually when the Tourism Satellite Account (TSA) is released

The TSA (produced by Stats NZ), is released once a year and gives a complete picture of national tourism spending. As well as releasing a provisional total for tourism spending in the year to March (2018), Stats NZ also revise spending totals for all other years, including the provisional total from the year before (2017). Following the release of the revised TSA totals, we adjust the MRTEs totals to match these.

The latest MRTE release included larger revisions due to a new TSA being released (it is released annually for the year-to-March data, the latest being for the year-to-March 2018). When estimating spend in the MRTEs (monthly figures), we project out from the latest TSA release (an annual figure). Therefore, in the October 2018 data, we were projecting out from the year-to-March 2017 TSA totals.

Because each region has a different spending profile, the revisions can affect each region differently.