Agricultural statistics

Lucy Darragh

We live in a world where the analysis of data influences almost every aspect of our lives. Yet, paradoxically, even as our reliance on data increases, official statistics sources are struggling under increasing pressure from budgetary constraints, shifted priorities and demands to modernise. This has particularly been the case for Australia’s agricultural statistics system, which has been eroded over the past decade.

For over a century, agricultural statistics produced by the Australian Bureau of Statistics (ABS) and the Australian Bureau of Agricultural and Resource Economics and Sciences (ABARES) have contributed to knowledge and understanding of the Australian agriculture sector. Agricultural policy, research and development activities and funding, and the production, marketing, trade and distribution of commodities have all relied on these statistics. An abundance of data and advances in big data analysis and visualisation has however recently been providing alternatives to some traditional agricultural statistics sources. For example, in the United States companies like Descartes Labs are providing crop forecasting services which are more timely and accurate than the United States Department of Agriculture forecasts, which have been the benchmark crop forecasting service for decades.

With the private sector disrupting many areas of official statistics collection, the relevance and responsiveness of public sector data is being challenged. It is important however that the public sector continues to produce robust and useful agricultural statistics. The private sector may provide alternatives where there is a commercial value proposition for the provision of information, however there will be many areas where statistics need to be provided as public good, including guiding appropriate policy and research priorities.

The crucial and underappreciated nature of robust and reliable agricultural statistics was underlined by the determination of the need for better agricultural statistics as one of the top priorities in the Australian Farm Institute’s (AFI) research priority-setting process. Analysis of agricultural data leads to discovery, informs predictions and forecasts, challenges assumptions, and is essential for data-based decision making. Perhaps most importantly though, is that in the absence of a robust public sector statistical system, the Australian agriculture sector faces inherent risks associated with policy deliberation and decision making based on unreliable or perception-based positions, rather than objective industry data.

There are a growing number of data sources relevant to Australian agriculture that could provide valuable supplements to the current survey-based national agricultural statistics collection. The ABS and ABARES are undertaking a program to understand the range of data sources available and determine their suitability for modernising the statistics system, and addressing some the flaws inherent in the existing system.

There is a sense of urgency in determining how the abundance of data that is being collected through digital platforms can be used to improve agricultural statistics systems. This sense of urgency is not confined to Australia – in fact a report from the US National Academies of Sciences, Engineering and Medicine has recently recommended how the US National Agricultural Statistics Service can improve its practices. The report looks at how complementary data sources, such as satellite and other remote sensing data and precision agriculture data, can be used to enhance the survey data already being collected in order to provide more precise county-level estimates of crops and farmland cash rents.

The AFI’s March 2013 report Is counting farmers harder than counting sheep? forecast some of the issues now at the forefront of concerns around the need to improve agricultural statistics. The report concluded that the Australian agricultural statistical system will continue to languish unless changes are implemented to provide clear leadership and to better integrate the system with the agricultural sector and the stakeholders it is designed to serve. Particularly relevant to contemporary issues is Recommendation 4 from the report which called for ‘the creation of a unique, interactive data warehouse’.

The AFI’s current project ‘Improving Agricultural Statistics’ will identify and evaluate agricultural and agriculture-related datasets that could be utilised to enhance Australia’s national agricultural statistics system, and to consider key barriers that will need to be overcome to enable their use.

The objectives of the research are:

  • A catalogue of potential sources of data that could be used to supplement or replace current national agricultural statistics.
  • A detailed understanding of the issues that would need to be considered in contemplating the inclusion of any of these in the national agricultural statistics system.
  • The identification of policies and industry initiatives that may assist in the incorporation of these alternative data sources in the national agricultural statistics system.

Image:  Sally Beech