Research and Development Survey 2025

Reference period The R&D Survey 2025 was posted out to respondents in August 2025, and was collected up until December 2025. The reference period for the latest survey was the 2024/2025 financial year. Businesses with balance dates falling between 1 January and 30 September supplied data for the year ending 2025. Businesses with balance dates falling between 1 October and 31 December supplied financial data for the year ending 2024. The table below details the balance dates respondents used for each survey. Balance date 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Proportion of responders (%) 31 December previous year 7 .. 8 9 8 8 8 8 9 9 31 March current year 44 .. 54 54 52 55 50 50 53 56 30 June current year .. 10 11 10 10 10 10 10 9 11 Other 11 .. 7 7 7 7 6 7 10 9 Not provided 27 .. 20 20 22 19 25 25 18 14 .. Data not collected Summary of changes to the R&D population In 2025, the R&D Survey was conducted only for the business sector, as a special iteration of the survey outside of its usual two-yearly cycle. During the usual cycle, government and higher education sectors are also surveyed. The sample for the R&D Survey: 2025 consisted of 4,998 units. The overall achieved response rate for the R&D Survey 2025 was 86 percent for businesses, the target response rate was 85 percent. We identified some businesses as key units - if their response to a survey question about R&D expenditure was considerable in the previous survey period. The target response rate was 100 percent for businesses identified as key units, and this was achieved in 2025. The sampling error on the Business sector R&D expenditure figure in 2025 was 5.5 percent at the 95 percent confidence level. The table below provides a numerical summary of the changes in the R&D population, as well as associated changes in the sample, number of respondents, and sampling errors. The population and sampling methodology has remained relatively unchanged since 2017. For more information on the population criteria and changes, see: Research and Development (R&D) Survey - Stats NZ DataInfo+ Sector 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Population Count Business 4,185 .. 4,512 4,775 5,192 5,382 5,688 5,465 5,561 5,436 Government 101 .. 97 .. 101 .. 87 .. 101 .. Higher Education 8 .. 8 .. 8 .. 8 .. 8 .. Total 4,294 .. 4,617 4,775 5,301 5,382 5,783 5,465 5,670 5,436 Sample Count Business 3,895 .. 4,159 4,386 4,767 4,948 5,228 5,023 5,112 4,998 Government 92 .. 90 .. 95 .. 87 .. 101 .. Higher Education 8 .. 8 .. 8 .. 8 .. 8 .. Total 3,995 .. 4,257 4,386 4,862 4,948 5,323 5,023 5,221 4,998 Number of respondents Count Business 2,978 .. 3,455 3,663 3,935 4,138 4,151 4,001 4,292 4,281 Government 80 .. 78 .. 80 .. 75 .. 81 .. Higher Education 8 .. 8 .. 6 .. 8 .. 8 .. Total 3,066 .. 3,541 3,663 4,021 4,138 4,234 4,001 4,381 4,281 Response rate(1) Percent Business 77 .. 83 84 83 84 79 80 84 86 Government 87 .. 87 .. 84 .. 86 .. 80 .. Higher Education 100 .. 100 .. 75 .. 88 .. 100 .. Total 83 .. 83 84 82 84 80 .. 84 86 Yes to R&D(2) Number Business 1,340 .. 1,642 1,814 1,908 1,960 1,862 1,825 1,922 1,926 Government 44 .. 46 .. 38 .. 39 .. 39 .. Higher Education 8 .. 8 .. 6 .. 8 .. 8 .. Total 1,392 .. 1,696 1,814 1,951 1,960 1,909 1,825 1,969 1,926 R&D performed as a percentage of respondents(3) Percentage Business 45 .. 48 50 48 47 45 46 45 45 Government 55 .. 59 .. 38 .. 51 .. 48 .. Higher Education 100 .. 100 .. 100 .. 100 .. 100 .. Total 45 .. 48 50 48 47 45 46 45 45 Sample errors(4) Percentage Business 6.3 .. 12.0 6.2 6.0 5.1 5.6 6.5 7.6 5.5 Government 3.6 .. 8.0 .. 3.1 .. 2.3 .. 2.8 .. Higher Education 0 .. 0 .. 0 .. 0 .. 0 .. Total 3.3 .. 6.7 .. 3.5 .. 3.3 .. 4.8 5.5 (1) Response rate = number of respondents divided by sample. (2) 'Yes to R&D' means that a business is carrying out and/or funding R&D. (3) R&D performers as percentage of respondents = 'yes to R&D' divided by number of respondents. (4) All years sample errors were updated in 2021, to handle variability of responses within the sample in accordance with current best practice. .. Data not collected. Note: Due to rounding some figures may not add to stated totals. All figures in this table are based off unweighted and unrounded counts. Response rates by industry The table below shows response rates by industry. The overall response rate was 86 percent, ranging from 83 percent to 90 percent. Industry 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Primary 74 .. 80 83 85 84 78 77 76 83 Manufacturing 80 .. 85 86 85 86 82 82 87 87 Food product manufacturing 78 .. 85 84 82 86 78 78 83 85 Beverage and tobacco manufacturing 74 .. 91 89 83 81 78 81 79 85 Textiles, clothing, footwear, and leather manufacturing 85 .. 86 85 85 86 89 86 91 90 Petroleum, coal, chemical, and associated product manufacturing 78 .. 85 84 86 85 80 86 84 85 Non-metallic mineral product manufacturing 81 .. 84 84 77 85 79 84 79 86 Metal product manufacturing 86 .. 90 90 87 89 88 85 93 90 Machinery and equipment manufacturing 82 .. 83 86 87 87 82 79 91 90 Other manufacturing 74 .. 83 88 86 88 87 82 88 86 Services 77 .. 83 83 81 83 79 79 84 85 Wholesale trade 78 .. 84 85 84 84 81 82 86 86 Scientific research and technical services 81 .. 84 79 85 84 80 82 85 87 Computer services 76 .. 83 83 77 81 77 76 81 82 Other services 75 .. 82 82 81 82 79 79 84 86 Total business sector 77 .. 83 84 83 84 79 80 84 86 Scientific research 92 .. 100 .. 100 .. 100 .. 100 .. Other government research 86 .. 85 .. 81 .. 84 .. 78 .. Total government sector 87 .. 87 .. 84 .. 86 .. 80 .. Total higher education sector 100 .. 100 .. 75 .. 88 .. 100 .. Total all sectors 77 .. 83 84 82 84 80 80 84 86 .. Data not collected Bioscience expenditure on R&D Based on consultation with key stakeholders, the questions that related to Bioscience expenditure on R&D was removed in 2016, as they were no longer relevant. Sampling error by industry The table below shows sampling error by published industry. Customers need to take into account these sample errors when using data at published industry level. All years sample errors were updated in 2021, to handle variability of responses within the sample in accordance with current best practice. 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Industry Sample error (percent) Primary 23.8 .. 49.8 30.2 29.7 26.0 42.8 40.8 33.7 30.7 Food product manufacturing 27.1 .. 17.8 22.7 12.6 18.2 10.5 13.9 75.6 29.5 Beverage and tobacco manufacturing 23.5 .. 57.7 32.7 63.9 13.9 10.8 12.5 12.4 12.2 Textiles, clothing, footwear, and leather manufacturing 14.2 .. 33.5 30.8 41.2 26.8 17.1 25.7 17.8 27.1 Petroleum, coal, chemical, and associated product manufacturing 14.5 .. 36.0 22.9 25.7 28.7 18.4 13.7 16.3 21.4 Non-metallic mineral product manufacturing 27.9 .. 54.7 93.8 121.6 17.3 32.6 37.2 13.4 12.7 Metal product manufacturing 27.9 .. 10.4 9.0 10.0 19.3 13.5 15.4 19.9 13.7 Machinery and equipment manufacturing 11.0 .. 13.1 12.6 11.7 10.0 8.2 8.1 9.5 7.8 Other manufacturing 21.5 .. 33.6 32.5 21.3 33.1 37.9 16.1 14.9 18.7 Wholesale trade 17.3 .. 40.3 16.6 21 17.0 10.5 13.8 11.0 29.0 Scientific research and technical services 18.8 .. 48.2 29.3 36.7 20.8 11.6 12.6 20.7 10 Computer services 12.5 .. 16.2 11.5 12 10.3 11.8 12.6 12.9 10.2 Other services 24.9 .. 45.3 17.5 11.7 11.2 15.4 20.7 17.9 14.7 Scientific research 0 .. 8.1 .. 0 .. 0 .. 0 .. Other government research 34.2 .. 36.4 .. 28.8 .. 20.3 .. 28.4 .. .. Data not collected Note: Due to rounding some figures may not add to stated totals. The survey sample was designed to minimise sample error at sector level. At published industry level, sample errors are higher, due to the smaller numbers in each of these groups. They quantify the variability that occurs by chance because a sample rather than an entire population is surveyed. For example, if the estimated R&D expenditure for businesses in the Primary industry in 2025 is $197 million, with a sampling error of 30.7 percent, this means there would be a 95 percent chance the true R&D expenditure for the Primary industry lies between $137 million and $258 million ($197 million plus or minus 30.7%). The sampling errors detailed above represent the variability for total R&D expenditure at the published industry level. Sample errors for other variables collected in the survey, as well as for other breakdowns, can be found in the downloadable csv as part of this release. en-NZ