The BOS is an annual three (sometimes four) part modular survey, which began in 2005. The first module is focussed on firm characteristics and performance. The second module alternates between biennial modules on innovation and business use of ICT. The third module is a contestable module that enables specific policy-relevant data to be collected on an ad hoc basis.

The BOS excludes businesses that, on the selection date: had fewer than six employees, had been in existence for less than one year and were in five specific industries. The survey is conducted using two-way stratified sampling, with stratification on rolling-mean-employment (RME) and two-digit industry according to the ANZSIC system.

The BOS is something approaching best practice in such surveys internationally. It has removed replication of surveys – and thus reduces respondent load and makes sampling simpler. It is explicitly designed with a panel element, which enables more sophisticated analysis to be undertaken allowing us to better understand issues of causality and – as the panel element increases – dynamic issues.
In common with many surveys conducted by Statistics New Zealand (SNZ) the survey is statutory and the front page of the BOS bears the imprimatur:

Open

Because of this, the BOS has a considerably higher response rate than comparable surveys internationally (for example, the 2004 Workplace Employment Relations Survey in the UK achieved a response rate of 64%).

Modules

The BOS is made up of three (and sometimes four) sections or modules.

Module A is always Business Operations

Module B alternates between Information and Communications Technology (even years) and Innovation (odd years).

Module C is a contestable module that enables specific policy-relevant data to be collected on an ad hoc basis. In practice, Module C (and sometimes D) has repeated topics such as Business Practices, International Engagement and Finance on a roughly four year rotation as well as one-off and ad hoc modules such as Health and Safety (2017), the Transition to a Low-Emissions Economy (2021), firm responses to COVID-19 (2020) and a range of skills and employment related topics. . Since the practices by which businesses are run is always going to be important for business and economic policy, and the fact that as a small, distance economy internationalisation will always be a topic of major importance, this rotation system is likely to remain.

One of the great benefits of the BOS is that most of the Modules have remained very stable over time, with only minor changes and additions to questions and their possible response items.  This stability has enabled consistent analysis of changes in firm behaviours over time (eg, Sanderson 2022), and the ability to look at causal relationships between activities and performance (eg, Fabling and Grimes 2021).  However, it has also led to some modules being quite dated, especially those dealing with rapidly changing topics such as Business use of ICT.  In 2021, Stats NZ and MBIE (the primary funder of the BOS survey) began an overarching review of the BOS, resulting in significant changes to the ICT module.  Other modules are being reviewed in 2022 and 2023 which may result in additional content changes, although these are expected to be less substantial as the remaining modules deal with topics which have seen less rapid change than ICT.

Questionnaires

The best place to start is with a questionnaire in your hand (or on your screen if you prefer). There are a number of reasons for this. First, it is always good research practice to see what the respondents were seeing when they were answering the questions. Second, you get the exact questions they were being asked, including whether it was a ‘tick one’ or a ‘tick all that apply’ question. Third, you see the routing - some people only get to answer a question if they answered yes or no to a previous question. Fourth, they include the item codes for questions that will make up part of the names in the data itself.

Open

Many questions have multiple parts and are represented by multiple variables. For example, ordered categorical variable like these:

If the respondent ticked “no difficulty” to whether they experienced difficulty recruiting managers and professional staff, there would be a variable called a3301_1 that would be equal to 1.

Useful things

The last information release on the 2022 BOS.

• https://www.stats.govt.nz/information-releases/business-operations-survey-2022/

Download time-series of aggregate data from Infoshare:

• https://infoshare.stats.govt.nz/?_ga=2.256915715.905089402.1696275799-1217382070.1686251669

Definitions and metadata from SNZ’s DataInfo+ site

• Methodology: Business Operations Survey | Stats NZ DataInfo+

• Definition of terms: Business Operations Survey | Stats NZ DataInfo+