Learning from the historical failure of farm management models to aid management practice. Part 2. Three systems approaches

Part 1 analysed the difficulties experienced in the field of academic Farm Management in making complex theoretical models relevant to farming. This paper highlights the important connections developed between the field of Farm Management economics and 3 key ‘systems’ ideas and tools in agricultural science in response to difficulties and opportunities. The first systems approach reviewed is the 20-year experiment by agricultural economists in using crop and animal production simulation models in management analyses. The second systems approach reviewed is Farming Systems Research (FSR), an approach characterised by on-farm experimentation with a management orientation. Many pioneers of FSR were Farm Management economists disenchanted by the inapplicability of economic theory to farm management. The FSR that emerged is interpreted as a scion of the early era of Farm Management prior to the coup by economics theorists in the 1940s. A third systems approach reviewed is a ‘soft’ intervention to facilitate farmer learning. Although evolving from FSR, this approach has surprising similarities to the ‘goal adjusting’ consulting performed by the legendary Australian Farm Management consultant cum academic, Jack Makeham. The paper concludes with discussion of a recent innovation that combines these 3 approaches. It uses a soft intervention approach that features farmers shaping their goals and expectations by ‘experimenting’ in a local, but virtual, environment provided by simulation of the production system using ‘hard’ models.

Learning from the historical failure of farm management models to aid management practice. Part 1. The rise and demise of theoretical models of farm economics

A potential source of lessons for agricultural modellers aspiring to influence farm decision making is the historical experience of agricultural economists in the field, variously termed ‘Farm Management Research’ or ‘Farm Management’. Although the histories of Farm Management in the USA and in Australia differ significantly, in both cases the field was originally characterised by pragmatic on-farm research by agricultural scientists and later taken over by agricultural economists committed to theory-based economic analysis to enable rational planning and decision making. But in both countries, it became painfully evident to reflective participants that model-based Farm Management was not proving relevant to practical managers of farms. An insightful few went further to conclude not just that theoretical models of practice had not been relevant but that they could not be relevant, and since the late 1970s, the field has been in crisis. In this series of 2 papers, we seek insights that might explain this extraordinary ‘market’ failure of models that generate theoretical best practice as a basis for intervention. As an ‘experiment’, the history of Farm Management is enriched by the discontinuity between 2 ‘eras’ characterised by 2 contrasting intervention approaches, an ‘early’ interactive and pragmatic era and a ‘late’ academic and theoretical era. In this first paper, after a brief history of the early pragmatic era and the ‘take-over’ by economic theorists, we analyse the ‘crisis of relevance’ that led to demise, relying heavily on the remarkable intellectual journey of John Dillon, the first Professor of Farm Management in Australia who turned from being elder economic theoretician to pioneer philosopher of pragmatic Farming Systems Research. The significant turn to Farming Systems Research by disillusioned Farm Management economists in the 1980s was preceded by a turn to another systems approach 2 decades earlier, that of agricultural systems modelling. Learning from the autecology of these significant systems efforts to influence the management of farms is the aim of the second paper in this series.

Trends in mortality rates for infectious and parasitic diseases in Australia: 1907-1997

AIMS

To characterize long-term mortality trends for infectious and parasitic diseases in Australia during the twentieth century, explore influencing factors and provide suggestions to health policy-makers.

METHODS

A descriptive study was conducted. Deaths due to communicable diseases from 1907 to 1997 were tallied, according to the International Classification of Diseases version 9 (ICD-9). Trends in infectious disease mortality in overall population and in the 0-4 years age group were examined and standardized by sex. Death rates were also studied for: (i) diarrhoea/enteritis, (ii) pneumonia and all respiratory diseases and (iii) tuberculosis.

RESULTS

There has been a substantial decline in -mortality from communicable diseases over the study period. The death rate dropped from 258.9 per 100,000 population in 1907 to 7.2 per 100,000 population in 1997. Six phases of the decline were observed.

CONCLUSIONS

A combination of improved living conditions and access to readily available treatments over the twentieth century played an important role in the reduction of infectious disease mortality in Australia.

Family self-medication and antibiotics abuse for children and juveniles in a Chinese city

To identify the determinants of self-medication and antibiotics abuse by parents treating their children aged between 2 and 18 over the previous year, an investigation was conducted in Hefei City, China in April, 1995. A total of 1596 students from a kindergarten, a primary school and a high school were included in the study, and 1459 completed questionnaires were collected (the response rate: 91.4%). The results showed the rate of parental self-medication for their children in the sample was 59.4%. It increased with children's age; about 51% of children had received parental self-medication on six or more occasions during the 1-year period and 32.8% on four to five occasions; there were associations between parental self-prescribers and sources of medicine and severity of disease. The rate of antibiotics abuse was 35.7%. Logistic regression analysis showed that there were significant associations between self-medication and payment of the mother's medical fees by employers, severity of diseases as well as the mother's educational level.

Seasonal rainfall variability, the incidence of hemorrhagic fever with renal syndrome, and prediction of the disease in low-lying areas of China

To investigate determinants of hemorrhagic fever with renal syndrome (HFRS) in low-lying areas of China, the authors studied Chuigang and Wanggang communities in Anhui Province. These adjacent farming communities have a population of about 100,000. Data were collected from the two communities in 1961-1977 and from Yingshang County in 1983-1995; information covered the incidence of HFRS, amount of precipitation, differences in the water level of the Huai River, density of Apodemus agrarius, autumn crop production, and areas of inundated farmland. Correlation and multiple linear regression analyses were used to estimate the relation between seasonal rainfall, density of mice, occupational factors, and occurrence of the disease. Associations were observed between the incidence of HFRS and the amount of precipitation, the water level of the Huai River, and the areas of inundated farmland in Chuigang community. The smaller the water-level difference, the less farmland was inundated and the higher the incidence of HFRS. In Wanggang community, the density of A. agrarius (r1=0.63, p=0.02), the water-level difference in the Huai River (r2=-0.81, p=0.007), and crop production (r3=0.96, p=0.005) were correlated with the incidence of HFRS. The regression analyses based on Wanggang community suggested that these indexes could be used as predictive variables, and the results from the model were well calibrated with the actual incidence of HFRS in that community (R2=0.88, p < 0.01) and Yingshang County (R2=0.91, p < 0.01).

Stochastic dominance and mean-standard deviation analysis: some critical issues

Stochastic dominance and mean-standard deviation analysis can be effective procedures for ranking risky alternatives that are expressed in terms of probability distributions of outcomes. However, the conditions applying to their use and their limitations need to be understood. These are set out in the paper, together with an extension that overcomes some constraints to the use of stochastic dominance analysis.