Australian Natural Resources Atlas

Natural Resource Topics

People - Adjustment - Background to the assessment of structural adjustment in agriculture

Background to the assessment of structural adjustment in agriculture

Data Sources

Population and Housing Census

The Australian Bureau of Statistics Population and Housing Census is conducted Australia-wide at five-yearly intervals, the most recent being in 1996; the project obtained data for this census and for those conducted in 1991 and 1986. These censuses provide information about individuals who describe themselves as farmers, and about their families. Data were aggregated to 1996 Statistical Local Area boundaries for each of the three censuses. Data was also purchased from these three censuses aggregated at Collector District level within each of the case study catchments.

Farm Census Commodity Items

During the period 1983 to 1997, the Australian Bureau of Statistics distributed an annual farm census form all Australian farming businesses meeting a minimum gross income criterion. All farm business operators are required by law to complete and return it. The Farm Census contains a series of questions on farm production items and management techniques. This project uses data from each of these farm censuses, aggregated to Statistical Local Area (SLA).

For Victorian case studies Australian Bureau of Statistics Farm Census data was available over this time period aggregated to Parishes. Parishes are a static entity, comprising part of the cadastral system for describing land title. Thus, parish data does not need to be concorded. However, parish data is still based upon changing Estimated Value of Agricultural Operations cut-offs and use of this data needs to take this into account. The equivalent cadastral unit in South Australia is the "Hundred", and farm census data is available in that state aggregated to the Hundred.

Some Important Definitions

What is a Statistical Local Area

A Statistical Local Area (SLA) is a geographic region used by the Australian Bureau of Statistics to report summarised aggregated data. Statistical Local Areas are based loosely upon Local Government Areas (Local Government Area). Local government areas generally wholly encompass two or more Statistical Local Areas. Statistical Local Area boundaries change in response to changing population distribution. Major changes in Victorian Local Government Areas boundaries in last decade have necessitated similar major changes to Statistical Local Area boundaries in that state. Indicators reported in this project have been presented aggregated at 1996 Statistical Local Area boundaries. Population and Housing Census data from previous census years has been re-aggregated to 1996 Statistical Local Area boundaries. Farm Census data has been concorded to 1996 Statistical Local Area boundaries. In this report indicators are presented mapped by Statistical Local Area at a scale which does not portray the status of some smaller rural Statistical Local Areas. An accompanying appendix report provides finer scale mapping which allows identification of all rural Statistical Local Areas (see Appendix figures 1-5).

Estimated Value of Agricultural Operations (EVAO)

The Australian Bureau of Statistics uses the data collected within its Farm Census to create a measure of farm size called Estimated Value of Agricultural Operations. Commodity values derived from a separate Farm Financial Survey are applied to produce reported in the Farm Census. The resulting constructed measure is a proxy for gross farm income. Estimated Value of Agricultural Operations are used as a basis for deciding whether to include or exclude agricultural businesses from aggregated Farm Census data, but are normally only reported in aggregate. We purchased tables of farm Estimated Value of Agricultural Operations distributions for Statistical Local Areas for the period 1983/84 to 1996/97. Break points within the classification were adjusted for inflation throughout this period.

Farm Establishment

The Farm Census data set has had an inconsistent structure over the period 1983-1997. Farm businesses (establishments) are included in data aggregation if the value of their production exceeds a minimum Estimated Value of Agricultural Operations. The minimum Estimated Value of Agricultural Operations required for inclusion within the census has varied inconsistently from $2,500 to $22,500. The National Land and Water Resources Audit purchased a partial set of Census data items was purchased for the period 1983/84 to 1996/97 based upon a consistent cut-off of $22,500. This data was then concorded to a consistent boundary set (1996 Statistical Local Areas) by the Bureau of Resource Sciences using a land use mask and apportioning items on a pro-rata basis according to area of agricultural land. This data set, whilst improving the utility of historical Australian Bureau of Statistics farm census data, is still limited because of its use of a fixed cut-off based upon nominal Estimated Value of Agricultural Operations. A second historical set of farm census data was purchased based upon an Estimated Value of Agricultural Operations cut-off specified in 1996 dollar form. This data set used an Estimated Value of Agricultural Operations cut-off of $30,000 ($1996). This subsidiary data set contained measures of the distribution of farm establishment numbers by indexed Estimated Value of Agricultural Operations categories. All counts of establishment numbers used in this report are based upon this data set.

Farmer

The categorisation of a person's occupation as farmer is based upon a self-description question used in the Australian Population and Housing Census. Respondents are asked their major occupation within the preceding week. The self-description of farmer is quite open to ambiguity.

Does the definition depend upon a respondent's dependence on farm income, his or her time allocation to off farm work, or a more fundamental self-image change? One of the possible changes in self-image, which could complicate any longitudinal analysis of data, is changing perceptions of women's role in the farming sector. We checked our data to determine if there had been any change in the relative propensity of women to describe themselves as farmers in NSW and Victoria. This analysis showed only minor changes in the proportion of NSW women describing themselves as farmers for the decade prior to 1996. There was no discernible change in Victoria. This result allows us to avoid correcting our time series for changes in the sex composition of the farm population.

A second possible variation in self-description of farmers may be due to changes in off farm work dependence. Whilst no direct study has been made of this matter, it is possible to deduce the extent of this influence by comparing the number of farm families in an area with the number of farm establishment. Geographic variation is significant. Farm establishment numbers significantly outnumber farm families on the fringes of major towns and in some other scattered locations. These are presumably areas where land prices are far greater than that which could be justified by agriculture. Landowners in these areas will be more likely to have significant off farm income sources. Seasonal off farm work is quite common in many agricultural industries. This means that the timing of the census will influence the count of farmers. The consistent timing of census administration over the past decade allows us to conclude that any error in the count of farmers will at least be consistent.

Farm Family

A farm family is defined as any family with a t least one member describing his or her major occupation as farming. Farm family numbers are significantly greater than farm establishment numbers in Statistical Local Areas along the spine of the Great Dividing Range. The causes of this phenomenon will be discussed in "Adjustment - strategies and outcomes since 1986". An initial speculation would be that these areas contain significant numbers of self-described farmers whose establishments do not meet the criteria for inclusion in the farm census. Farm family and farm establishment numbers are closer to equality on the plains inland of the Great Dividing Range. These geographical variations in self-description must be considered in any interpretations of subsequent analyses. Data sets used for agricultural structural adjustment analysis

Table 1 Description of data used in this project

Farmers*
Definition Time series Data
Individuals who described their major occupation as farming 1986,1991,1996 Age
Income
Migration
Education
Farm Families*
Definition Time series Data
Families with at least one member who described his or her major occupation as farming 1986,1991,1996 Income
Farm establishments*
Definition Time series Data
Farm establishments with a EVAO of $30,000 in 1996 dollars. ie the cut-off has been indexed to account for inflation. 1986,1991,1996,1997 Selection of major commodity areas and counts
Farm establishments with a nominal EVAO of $22,500. Inflation not factored into the cut off. 1997 Selection of major commodity areas.
Farm establishments with variable cut offs 1984-1997 Full farm census data
Farm establishments with variable cut offs 1984-1997 Distribution of farm EVAOs using EVAO category boundaries indexed for inflation.

Data source = ABS Population and Housing Census: Customised Tables

Seasonal conditions and market prices in the census years

The key years in which the data used in this study was collected were 1985-86 1990-91 and 1995-96. Commodity prices and seasonal conditions during the inter-censal period will be one of the factors driving structural change. Seasonal conditions and commodity prices during the last year of the inter-censal period will have a major influence upon farm and family income data reported to the census.

1981-96: The period leading up to the first census included in this analysis was characterised by generally declining prices for both cropping and grazing industries. The exception was the beef industry which experienced an improvement in real prices. The middle of this period coincided with a severe el Nino event which resulted in low rainfall or drought across much of eastern Australia. During the 12 months preceding the 1986 Agricultural Census rainfall was average across most of the Australian agricultural zone (QDPI & QDNR 2001). Major rainfall deficits were experienced in central Australia, the Gulf country and coastal South Australia. In these areas production will have been depressed (view maps below).

1986-91: This is the first inter-censal period in our study. The most obvious feature of this period was a major peak in wool and wine grape prices. Most cropping commodity prices showed little overall change over this inter-censal period. There was a significant decline in wheat prices and a significant rise in canola prices in the final year. Adjustment processes would reflect a period of stable prices in most cropping industries, beef, sheep meat and dairy industries. Adjustment processes in the wool industry would reflect a period of buoyancy, though income data would reflect the onset of low prices at the end of the inter-censal period. Climatic conditions were varied across Australia. During the first two years of the period low rainfall was experienced in south west Australia, coastal Queensland, northern Tasmania and Gippsland. Central Queensland and the Eyre Peninsula experienced rainfall deficiencies in 1988. The 12 months preceding the 1991 census saw low rainfall in the Great Southern of West Australia and much of south east Queensland. Estimated Value of Agricultural Operations data from these regions will reflect these climatic conditions with reduced production in both these areas.

1991-96: This inter-censal period was dominated by the period of prolonged low wool prices and a drought in the central NSW and southern Queensland wheat belt. Parts of the Darling Downs experienced only one year of average or above average rainfall. The year of the 1996 census saw average rainfall across much of Australia, with above average rainfall in the Darling Downs. Rainfall deficit was experienced in the Eyre Peninsula of South Australia. There is remarkable similarity in the climatic conditions prevailing in the years of the 1986 and 1996 censuses. The breaking of the Queensland drought and a rise in grain prices inflated reported farm family incomes well above that received during much of the inter-censal period. Adjustment processes measured in this report reflect the duration of both these difficult conditions during much of the period. In contrast, wine grape prices rose significantly during the same period. Adjustment patterns in grape growing horticultural districts will reflect these buoyant circumstances.

Annual rainfall deficits for 1986,1991 and 1996 (Source: QDPI/DNR Long Paddock Web Site):
View map: Rainfall - relative to historical records: 12 months up to and including June 1986
View map: Rainfall - relative to historical records: 12 months up to and including June 1991
View map: Rainfall - relative to historical records: 12 months up to and including June 1996

Before you download

Most publications are downloadable as PDF files. Adobe Acrobat Reader  is required to view PDF files.

If you are unable to access a publication, please contact us to organise a suitable alternative format.

Key

   Links to an another web site
   Opens a pop-up window