australian growth industry factors influencing productivity growth philip kokic, alistair davidson and veronica boero rodriguez
Growth in total factor productivity is continuing to ease in Australia’s grains industry.
Moisture availability is the main driver of productivity on farms.
Land use intensity is another major influence, with its impact dependent largely on natural soil fertility and management practices.
Other factors that have a significant impact on farm productivity in some or all regions include access to finance, land area, education of farmers, crop specialisation, investment income and corporate ownership.
As moisture availability is largely beyond the control of farmers, its impact has been netted out of the analysis presented here.
With the effect of moisture availability removed, the Australian grains industry is estimated to have realised an underlying productivity improvement of around 2.6 per cent a year on average over the sixteen years from 1988-89 to 2003-04 — considerably less than the rate of 3.8 per cent over the seventeen years to 1993-94 estimated in a previous ABARE study.
productivity growth critical to farmers
Sustained productivity improvements have long been the engine of growth of Australia’s agriculture sector. Farmers have had to pursue more efficient ways of producing more output from less input to offset declining ‘terms of trade’ and maintain viability. Over time, farmers’ terms of trade have fallen, on average, as the prices they have received for their outputs have fallen relative to the prices they have paid for their inputs.
Strategies employed by farmers for making productivity gains in the past have depended on access to a broad range of inputs, both ‘market’ (purchased) and ‘nonmarket’ (nonpurchased). However, there is a growing realisation that the environment in which farms operate in the future could alter farmers’ access to some of these inputs, particularly the nonmarket inputs. Factors such as climate variability and climate change, vegetation regulation and rural migration could also have an impact on the effectiveness of using past strategies to make productivity gains in the future.
Total factor productivity analysis has been used as a framework for understanding the forces that affect productivity. Past studies of productivity in the grains industry have typically focused on growth in industry productivity over time (Knopke, O’Donnell and Shepherd 2000). However, as those studies relied on data aggregated at an industry level, the scope to investigate the factors influencing productivity changes at an individual farm level has been limited.
In a recent ABARE study (Alexander and Kokic 2005), productivity estimates were produced at the farm level for two different years. This allowed factors that were important in explaining differences between farms to be investigated. However, it was not possible to adequately identify factors that had a significant impact on productivity over time. Indeed, any changes in weather patterns or restrictions on land use could be expected to have a major influence on farm productivity both over time and between regions. When measuring total factor productivity, what is not well known is the relative importance of individual agricultural inputs used by growers from the diverse range available.
This article summarises the main findings of ABARE’s latest research on productivity growth in Australia’s grains industry (Kokic, Davidson and Boero Rodriguez 2006). In that research, an economic framework for systematically investigating a broad range of factors that potentially influence productivity growth in the grains industry was used. The framework is based on a rural livelihoods approach (Ellis 2000) that is founded on a diverse range of market and nonmarket capital stocks (physical, natural, financial, human and social) as well as a broad range of factors that modify farmers’ access to these agricultural business inputs. The statistical approach used in Kokic et al. (2006) to evaluate factors assembled in the rural livelihood productivity model provides new insights into the factors that influence productivity growth in the grains industry.
identifying factors that influence productivity growth
The factors likely to influence productivity growth on Australian grain farms are also likely to be the factors that can influence growth in rural incomes. The rural livelihoods framework (Ellis 2000; Nelson et al. 2005) provides a convenient approach to systematically identifying such factors, as shown in figure A.
Rural livelihood strategies are built on a set of five asset classes or stocks of capital from which farmers are able to undertake production. Farms with access to all these dimensions of capital are likely to be more resilient to external changes and shocks, and consequently better able to remain viable during difficult times. They may also be better placed to improve productivity over time. Furthermore, sustainable productivity growth may depend on the range of capital types to which farmers have access and the degree of substitutability between these different types of capital stock. The five types of assets or capital stocks identified by Ellis (2000) are:
physical capital: assets brought into existence by economic production processes (for example, tools, machines, infrastructure and land improvements)
natural capital: the natural resource base that yields products used by human populations (for example, land, water and vegetation)
human capital: factors that influence the productivity of labor (for example, education, skills, health and management capacity)
financial capital: stocks of cash and other financial assets that can be used to purchase either production or consumption goods, and access to credit
social capital: the social norms, networks and trust that facilitate cooperation within or between groups.
Access to the five types of capital occurs through processes that are largely outside an individual farmer’s control. It is therefore likely that substitution between asset classes is affected by imperfect information, transaction costs, externalities and a range of other factors facing farm business units. These include the numerous social, economic and policy (government and industry) considerations that affect a farmer’s access to the capital stocks used in income generating activities (known as livelihood strategies).
Ellis (2000) identified five types of mediating processes or modifiers that affect access to asset/capital stocks (figure A):
social modifiers: factors, including age and possibly gender, that may inhibit capabilities and choices or enhance access to the five asset classes
institutional modifiers: formal rules or conventions that constrain human interactions, including legal systems that affect private property rights and the way in which markets operate
organisational modifiers: groups, including farmer organisations and government agencies, that are bound by a common purpose
trend modifiers: time related changes, including relative price trends, rural migration and technological improvements, that individuals have no control over
shock modifiers: periodic destruction or loss of capital stocks caused essentially by random events, including drought, disease and pest incursions as well as unsecured losses incurred following corporate bankruptcies.
factors influencing productivity
Factors from four (physical, natural, human and financial) of the five asset classes were found to have a significant influence on productivity growth in most regions, according to the analysis reported here. It was not possible to draw conclusions on the importance of social capital in determining productivity growth.
moisture availability is the dominant influence on productivity
The moisture availability index used in the current analysis was obtained from the Queensland Department of Primary Industries (Potgieter, Hammer and Butler 2002). It measures the amount of soil moisture available for wheat production during the winter growing season.
Moisture availability was found to be the dominant factor affecting total factor productivity at any point in time. The effect of moisture was considered so dominant and largely beyond farmers’ control that all further analysis was done excluding the effects of moisture availability in order to develop a clear understanding of the technical efficiency with which grain producers combine inputs to produce outputs. However, the efficiency with which available moisture is used by farmers can be affected by cropping practices. For example, direct drill and minimal till are likely to improve total factor productivity in the southern region during drier years (map 1, which shows regions defined by the Grains Research and Development Corporation).
impact of land use intensity and other factors
Land use intensity was the other factor (apart from moisture availability) that has a substantial impact on total factor productivity. Its impact largely depends on natural soil fertility and management practices (for example, double cropping). Other factors that have a significant impact on total factor productivity at the farm level in some or all regions include access to finance, farm land area, farmers’ education, crop specialisation, investment income and corporate ownership.
Land degradation in the form of depreciated natural capital through wind erosion, water erosion, dryland salinity, soil acidity, soil sodicity and loss of soil structure did not have a significant effect on total factor productivity in any region according to the analysis undertaken in this study. However, it is possible that variations in land degradation effects could have been picked up in other variables analysed.
results vary across farms and regions
Considerable variation remains in productivity growth between and within farms that is not explained by factors explicitly included in the analysis (such as moisture availability, land area, land use intensity and crop specialisation). This variation is likely to have a corresponding impact on income distributions. Productivity was found to be increasing significantly as a consequence of factors not included in the analysis — by 1.5 per cent a year in the northern region, 1.0 per cent in the southern region and 1.8 per cent in the western region.
productivity growth trends
Total factor productivity (TFP) is defined as the ratio of a quantity index of all marketable outputs to the corresponding quantity index of all marketable inputs (Coelli, Rao and Battese 1998). It is a measure — over a data period — of the annual proportional rate of improvement in the technical efficiency with which farmers combine marketable inputs to produce marketable outputs.
In Australia’s grains industry, total factor productivity increased at an average growth rate of 1.86 per cent a year over the sixteen years from 1988-89 to 2003-04 (table 1). For this study, Australian grain producers have been defined as farmers in the specialist grains industry and the mixed livestock–crops industry. A further breakdown of the growth rates into separate GRDC regions is given in table 1.
An earlier study by Knopke et al. (2000, p. 13) indicated that productivity growth rates realised by specialist crop producers (3.6 per cent) were noticeably higher than the average rate achieved by mixed livestock–crops producers (2.6 per cent) over the period 1977-78 to 1998-99. However, the results from the present study suggest that the gap has narrowed and that productivity growth rates for specialist crop farms and mixed livestock–crops farms are now roughly equal, at 1.8 per cent and 1.9 per cent respectively.
The all regions TFP growth rate of 1.86 per cent estimated in this study is considerably less than the rate of 3.8 per cent obtained by Knopke et al. (2000, p. 15) over the period 1977-78 to 1993-94. In that study, a decline in productivity growth rates in the mid to late 1990s was observed. In this study, the all regions estimate provides further evidence that total factor productivity growth in Australia’s grains industry has continued to decline in recent years.
Part of the reason that the TFP growth rates presented here are lower than those reported by Knopke et al. (2000) is because of the severe drought toward the end of the sixteen year analysis period. Moisture availability effects are the largest single factor affecting TFP and this factor is beyond farmers’ control. It is therefore appropriate to measure the underlying TFP growth rate excluding the effect of moisture availability (table 1) to better understand how efficiently farmers are combining market inputs to produce market outputs.
Even after allowing for the effects of moisture availability, including the 2002-03 drought, the TFP growth rate adjusted for moisture availability at the national level of 2.58 per cent remains less than the growth rate of 3.8 per cent obtained by Knopke et al. for the period 1977-78 to 1993-94. This suggests that productivity growth rates for farms in the grains industry may be declining.
The productivity growth rates referred to so far, and presented in the first two columns of table 1 and also those produced previously by ABARE (Knopke et al. 2000), are for a given industry or aggregate region as a whole. These estimates are useful for comparison with other sectors of the economy or for international comparisons where individual farm level data are not available or where productivity estimates have been estimated using incompatible methodologies. However, such aggregate growth estimates conceal the variation in productivity growth occurring on individual farms.
1 growth in total factor productivity, 1988-89 to 2003-04
excluding moisture effect
region
region
farm average
growth rate
growth rate
growth rate
%
%
%
GRDC region
northern
0.82
1.26
2.03
*
southern
2.2
*
2.81
*
2.03
*
western
1.8
2.67
*
2.47
*
all regions
1.86
*
2.58
*
2.19
*
* Estimate is statistically significant at the 5 per cent level.
farm average growth rates
Any industry or particular region is typically dominated by a small proportion of large farms. Accordingly, the growth rates computed at an aggregate level will underweight the contribution of smaller enterprises. Thus to reflect a more accurate picture for all producers In the grains industry, farm average productivity growth rates were produced for this study.
The farm average TFP growth rate, excluding the effect of moisture availability, of Australian grain producers from 1988-89 to 2003-04 is 2.19 per cent, considerably less than the industry aggregate growth rate of 2.58 per cent in table 1. This suggests that the TFP growth rate of smaller producers in the grains industry was less than that of larger producers over the sixteen year period. The farm average TFP growth rate (excluding the effect of moisture availability) was positive in all three GRDC regions (northern, southern and western) in the sixteen year period from 1988-89 to 2003-04. TFP growth rates in the western region were the largest over the sixteen years to 2003-04.
region average growth rates – key components
The key components of total factor productivity have been separated and reported in table 2 in order to identify where the main changes are occurring. On average, total annual output from cropping and mixed livestock–crops farms grew in all three GRDC regions over the sixteen years to 2003-04: western region (4.8 per cent), southern region (3.8 per cent) and northern region (1.1 per cent). The strongest output growth occurred in the western region, largely because of the significant increase in cropping and decline in wool production. In 1988-89, grain and wool production in the western region accounted for around 60 per cent and 28 per cent of total output respectively. In 2003-04, however, these proportions had shifted to 74 per cent and 9 per cent respectively.
The shift from wool to cropping was most dramatic in the southern region, where average farm wool output declined at 13 per cent a year over the sixteen years, to the extent that wool’s share of average farm output was around 6 per cent in 2003-04. Wool production in the northern region trended in the same direction. All three GRDC regions exhibited significant increases in income from other nonagricultural sources.
The quantity of total inputs used on grain properties in the northern region remained roughly constant between 1988-89 and 2003-04. However, there was a significant decline in capital, fuel and labor inputs, concurrent with a significant increase in consumption of chemicals. Increased chemical use is consistent with greater adoption of low tillage crop husbandry practices that can deliver higher yields through greater moisture retention. The decline in the level of farm capital inputs used on grain farms is largely explained by the decline in buildings and other farm improvements as well as the reduction in the sheep flock.
In contrast to the pattern in the northern region, there was significant average annual growth in total input use in the southern region (1.8 per cent) and the western region (2.3 per cent) over the sixteen years to 2003-04 (table 2). The growth in the quantities of inputs used was largely a consequence of the growth in material and services used.
In 2003-04, these inputs accounted for roughly half of all inputs used in both of these regions. A significant proportion of the growth in total inputs used also resulted from a substantial increase in the use of crop chemicals — 9 per cent a year in the southern region and 18 per cent a year in the western region. Crop chemicals accounted for at least 10 per cent of all input use in all regions in 2003-04 — more than the farm average input share for fuel in all regions. At the same time, there was a slight reduction in capital inputs, consistent with the greater adoption of minimum tillage techniques. The growth in crop chemical use was largest in the western region, which also had the highest adoption of direct drill and minimum tillage (83 per cent), compared with the northern region (58 per cent) and southern region (61 per cent) in 2001-02.
2farm average growth rates (excluding moisture effect), 1988-89 to 2003-04, and share of input or output costs in 2003-04
cropping and mixed livestock–crops industries, by GRDC region
northern
southern
western
growth
share
growth
share
growth
share
%
%
%
%
%
%
total factor
productivity
2
*
2
*
2.5
*
all outputs
1.1
100
3.8
*
100
4.8
*
100
crops
1.3
61
3.7
*
73
4.7
*
74
livestock
2.6
28
–2.9
*
16
0.6
14
wool
–6.1
*
5
–13.1
*
6
–2.9
9
other
7.9
*
5
7.3
*
5
10.7
*
4
all inputs
–0.9
100
1.8
*
100
2.3
*
100
land
–0.9
11
1.2
*
10
2.7
*
9
capital
–3.8
*
25
–0.8
*
22
–0.1
22
labor
–1.7
*
17
0.4
20
0.1
17
all materials
0.4
23
3.4
*
21
5.9
*
26
– crop chemicals
6.1
*
11
9
*
10
17.7
*
14
– fuel
–1.9
*
7
1.3
*
8
3.5
*
7
all services
–1.1
24
2.1
*
26
2.6
*
26
– contracts
–7.6
4
0.6
4
0.4
3
* Estimate is statistically significant at the 5% level.
zone average growth rates
Even at a regional level, farm average TFP growth rates conceal marked subregional performance variation, reinforcing the need to disaggregate further to the level of GRDC agroecological zones (table 3, map 2). The average farm TFP growth rates (excluding the effect of moisture availability) ranged from 0.68 per cent in Queensland’s central zone (although not significantly different from zero) to 3.36 per cent in the Victorian high rainfall zone. Productivity growth rates in the GRDC agroecological zones also vary in their degree of sensitivity to moisture availability as shown in map 2 (data in table 3).
3farm average TFP growth rates (excluding moisture effect) and sensitivity to moisture availability, 1988-89 to 2003-04
cropping and mixed livestock–crops industries, by GRDC agroecological zone
GRDC agroecological zone
growth rate excl. moisture effect
sensitivity to moisture availability
%
%
northern region
Queensland central
0.68
0.96
*
New South Wales north east and Queensland south east
1.7
*
0.49
*
New South Wales north west and Queensland south west
3.3
*
1.47
*
southern region
New South Wales and Victoria slopes
2.87
*
0.83
*
New South Wales central
1.06
0.9
*
Victoria high rainfall
3.36
*
0.59
South Australia Victoria mallee
2.57
*
0.92
*
South Australia Victoria Bordertown–Wimmera
2.25
*
1.27
*
South Australia midnorth – Lower Yorke, Eyre
1.54
*
0.58
*
western region
Western Australia mallee and sandplain
2.28
*
1.62
*
Western Australia eastern
0.89
1.09
*
Western Australia central
2.6
*
0.78
*
Western Australia northern
3.13
*
1.12
*
* Estimate is statistically significant at the 5 per cent level.
It is unclear whether there is a direct relationship between productivity growth and sensitivity to moisture availability. However, it is evident from comparing the two panels of map 2 that some agroecological zones that are highly sensitive to moisture availability also have higher than average farm TFP growth rates (excluding the effect of moisture availability) — in particular, the New South Wales North West and Queensland South West zone in the northern region and Western Australia’s northern zone.
conclusion
The research undertaken in this study has revealed that the factors likely to have a significant influence on individual farm productivity performance are more diverse than has been reported in previous research. The findings also support Gollop and Swinand’s (1998) claim that total factor productivity is a biased measure of how well producers are allocating scarce resources because it fails to take into account changes in both nonmarket inputs and outputs. Taking account of nonmarket agricultural inputs and outputs produces a better measure to capture the technical efficiency with which producers combine all market and nonmarket inputs to produce all market and nonmarket outputs.
Considerable variation in productivity growth was found between and within farms that was not explained by factors explicitly incorporated in the analysis. The extent to which such productivity variations are a consequence of the management decisions made by individual farmers may have important consequences for the design of government policies that are intended to address fluctuations in climatic factors.
references
Alexander, F. and Kokic, P. 2005, Productivity in the Australian Grains Industry, ABARE eReport 05.3, Canberra (www.abare.gov.au).
Coelli, T., Rao, D. and Battese, G. E. 1998, An Introduction to Efficiency and Productivity Analysis, Kluwer Academic Publishers, London.
Ellis, F. 2000, Rural Livelihoods and Diversity in Developing Countries, Oxford University Press, England.
Gollop, F.M. and Swinand, G.P. 1998, ‘From total factor to total resource productivity: an application to agriculture’, American Journal of Agricultural Economics, vol. 80, pp. 577–83.
Knopke, P., O’Donnell, V. and Shepherd, A. 2000, Productivity Growth in the Australian Grains Industry, ABARE Research Report 2000.1, Canberra.
Kokic, P., Davidson, A. and Boero Rodriguez, V. 2006, Australian Grains Industry: Factors Influencing Productivity Growth, ABARE Research Report 06.22 Prepared for the Grains Research and Development Corporation, Canberra, November.
Nelson, R., Kokic, P., Elliston, L. and King, J.-A. 2005, ‘Structural adjustment: a vulnerability index for Australian broadacre agriculture’, Australian Commodities, vol. 12, no. 1, March quarter, pp. 171–9.
Potgieter, A., Hammer, G. and Butler, D. 2002, ‘Spatial and temporal patterns in Australian wheat yield and their relationship with ENSO’, Australian Journal of Agricultural Resources, vol. 53, pp. 77–89.
