Use of the Statistics New Zealand Business Register

Use of the Statistics New Zealand Business Register for the agriculture industry and the not for profit sector Geoff Mead geoff. mead@stats. govt. nz

Introduction • Use of the Statistics NZ Business Frame (BF) for – Agriculture production statistics – Non profit institutions satellite account • Challenges • Benefits • Future developments

Statistics NZ Business Frame (BF) • Based on administrative tax data • All enterprises over a minimum threshold of tax activity • 460, 000 enterprises

Agriculture industry in NZ • Temperate climate suited to farming activities • Major industry, over 44% of exports • 32. 4 mill sheep • 5. 9 mill dairy cattle • 1. 6 mill ha plantation forestry • 33, 000 ha grapes • Formal businesses

Agriculture production statistics • Collected since 1861 • Annual agriculture production census from 1956 to 1987>> Census maintained the frame of farms, frame was essential for the census frame list of farms annual census annual etc census frame list of farms

• From 1988 to 2001 – financial constraints • could no longer afford the annual census • list of farms was not maintained – mix of censuses, sample surveys, no surveys – a number of different frames used, none entirely satisfactory • • • the traditional list of farms, only partially maintained list of farms integrated with BF industry lists other administrative land based sources no comprehensive survey feedback to the BF

From 2002……. . • From 2002 in partnership with the Ministry of Agriculture and Forestry – new agriculture production statistics programme – 5 yearly censuses, sample surveys in other years – resourced • Statistics NZ Business Frame (BF) used as the frame

Why was the BF used from 2002? –BF has comprehensive coverage of all enterprises with over a minimum threshold of tax activity –Benefits in coherence and consistency with other statistical outputs –BF existed, low cost compared with other options –No viable alternative administrative data based frame –Knowledge of the weaknesses of using the BF for agriculture production statistics

• Robust agriculture production statistics are being produced using the BF • Mainly a postal collection • Processes to update the BF with agriculture survey feedback – supplements admin data updates to BF – improved BF quality

• Sheep • Dairy cattle

Comparison of satellite imagery to agriculture production statistics based on BF Type of land (Land Cover Database categories) Satellite imagery 2002 Agriculture Production (Land Cover Database) survey 2002 ha ha High producing grassland 8, 886, 000 8, 243, 000 Exotic (plantation) forest 1, 962, 000 1, 828, 000 Primarily horticulture (includes horticulture, viticulture, cropping) 417, 000 534, 000

Coherence and consistency with other statistical outputs

BF coverage • Missing very small informal farms outside tax system • • Hobby farms, lifestyle farms Common in New Zealand on fringes of urban areas Owner works in town No solution to date • Quality caveats for users on coverage

Challenge one farm many enterprises multiple survey reporting units double counting risk Enterprise 1 farm operator Enterprise 2 land owner Managed by • Correct industry codes on BF • Filter questions in survey questionnaire • Flags on BF, share farmers etc Enterprise 3 share farmer

Feedback from the agriculture production survey to the BF BF population of enterprises in agriculture & forestry 90, 000 • Improved BF quality – Industry classification updates 80, 000 70, 000 60, 000 50, 000 40, 000 30, 000 20, 000 10, 000 0 2001 2002 2003 2004 2005 2006 2007 2008 2009

Future developments BF and agriculture production statistics Ideal – integration with: 1. land based administrative systems 2. administrative systems that record livestock for disease control Incremental steps towards the ideal, but unlikely to be fully achieved in the foreseeable future Land admin data BF tax admin data Livestock register admin data

A population of not for profit units • Statistics NZ published a non profit institutions satellite account in 2007 • BF was the foundation of the population

Additional administrative data sources used to extend BF coverage • BF coverage based on VAT and employment tax sources – Market orientated – And employers • Many small units outside the BF coverage – In tax system but below BF tax thresholds or – Outside tax system

BF coverage extended with administrative registers of 1. Incorporated societies 2. Charitable trusts Care required to avoid duplications Charitable Trusts Incorporated societies BF

Number of non profit institutions • 97, 000 non profit institutions identified – 18% from BF – 82% from other registers – BF covered all employers and units with significant turnover

Quality of population • Medium sized town used • Gathered all possible lists of non profit institutions • Matched lists against population constructed from BF and other admin registers • Lists had an additional 8% of units

Conclusion • Two outcomes – BF provides a robust population for agriculture production surveys – BF provides the core but requires extension for non profit institutions • Benefits in using centralised business register for as many outputs as possible – Coherence across statistical outputs – Efficiency – BF quality • Future developments – Extending BF beyond market orientated coverage – Ideally integration with other administrative sources of agriculture data