Abstract
Keywords
Rural suicide rates have increased significantly over recent decades in comparison to metropolitan areas [1], particularly in youth and in more remote and less populated areas [2–4]. Urban-rural differentials have also been noted more generally in other mortality and morbidity data [5]. Other more specific geographical correlates point to more complex trends in suicide with very high rates in some areas, but very low rates in other areas [6, 7]. A number of proposed factors to explain these patterns include differences in personal vulnerability and coping behaviours, isolation, lack of access to services, changes to the family unit and wider structural change associated with the rural economy [8].
This paper investigates the rural subpopulation of the farming community in the context of these urban-rural suicide differentials. Comprehensive data in Australia is currently lacking. Gun et al. [9] investigated suicide in Australian occupational groups for Australia and South Australia. This study reported a rate per 100 000 for Australian male farm managers and agricultural labourers of 47.7 and 54.6, respectively, for the year 1993. The Australian Bureau of Statistics (ABS) reports that nationally the male rates for 1992 and 1994 were 21.1 and 20.7, respectively [10]. Data for females was also reported for these groups, however frequencies were very small, and the rate unstable [9]. A number of North American and European studies have noted an elevated rate of suicide in male farmers and farm workers [11–15], and that rates differ across occupational groups [16]. Male farm suicide rates were higher than those of other rural males [13], differed across the farm work classifications of ‘farmer’ and ‘farm worker’ [14], and remained relatively stable across geographical areas [12]. The method most commonly associated with these farm suicides was firearms [11, 15]. This method has also been the most common in rural Australian areas. Other common methods of farm suicide included hanging and carbon monoxide poisoning [11], and hanging rates have increased in recent years [3].
In contrast, the hypothesized effects of changes to the rural economy and industry restructuring in Australia referred to above [8] were not reflected in investigations of farm suicide rates in Canada [17]. Nor were the elevated rates reported elsewhere. Pickett et al. [17] estimated the overall farm suicide rate to be 7.2 per 100 000, with regression analyses failing to find associations based on economic indicators. Similar studies of farm suicides reported that male suicide rates were similar to the national male rate of suicide [18, 19]. While a number of negative risk factors were identified associated with labour organization, the seasonal attributes of work, and to a lesser extent exposure to herbicides and insecticides, the protective factors associated with social support traditionally available were cited as responsible for the low farm suicide rate [18, 19]. Other positive factors referred to included marriage, having more than one resident on a farm, and a higher level of education [18].
It is suggested that these trends may not relate to the Australian farming population. There are differences in agricultural production systems and economic subsidization which arguably change the nature of occupational exposure between North American and European agriculture and Australian agriculture. Accordingly, this study aims to identify Australian farm suicides for the period 1988–97, and investigate their underlying patterns and trends. This will predominantly be based on crude frequency data; however, estimated age standardized rates will be generated for comparison to national rates. Preliminary exploration of farm suicides in terms of general agricultural economic conditions will also be made.
Method
Suicide data for the period 1988–1997 were obtained from the ABS. Occupations of interest were defined for two groups – farm managers and agricultural labourers. Occupational information was coded based on the first edition of the Australian Standard Classification of Occupation (ASCO) obtained from the ABS for data from 1990 onwards using the codes ‘1400 – Farmers, and farmer managers’ and ‘8200 – Agricultural labourers and related workers’. Prior to 1990 the Demography of Occupations system (DEMOSS) was used. The equivalent codes were sought and included ‘0210 – Farmers and graziers, farm managers’ and ‘0260 – Hunters and trappers, farm workers including farm foremen’. The classifications pertaining to shearers were not included as preliminary analyses revealed that frequencies were very low. It was also expected that using these occupational classifications the number of female suicides identified would be an underestimate of the actual frequency of female suicide on Australian farms. It is suggested that in many cases females on farms would not consider themselves as either a farm manager or labourer, even though their role may involve similar duties and occupational exposure. Frequencies in terms of age, sex, and method were determined for this subset of Australian suicides.
Currently, an accurate reference population for the farm labour force is not available. An approximation of this population was obtained based on linear interpolation of ABS labour force statistics for each year and farm residence estimates as outlined in Garnaut and Lim-Applegate [20]. Crude rates based only on ABS labour force population estimates for the study period were generated to establish preliminary rates. The age and sex structure for the labour force data used is not available for years prior to 1995, however, estimates of this age-sex structure have been generated by Garnaut and Lim-Applegate [20] for the years 1990–91 and 1994–95. The estimates defined by Garnaut and Lim-Applegate [20] are based on survey data from a proportion of Australian farms. This population data estimates the farm resident population and not the labour force. The age-sex structure reported by Garnaut and Lim-Applegate [20] was applied to the ABS labour force estimates for age groups of interest. This required breaking estimates into five-year age groups and using linear interpolation and extrapolation to establish estimates for the intervening years between 1990 and 1995. Age-adjusted rates per 100 000 were then calculated. The reference population used in the analyses was based on the national population obtained from the most recent census [21]. National rates have been taken from reported ABS data [21].
Analysis of farm suicides in relation to farmers’ terms of trade was also completed. Farmers’ terms of trade is a ratio of the index of prices received by farmers to the index of prices paid by farmers [22]. This ratio is an indication of the relative ‘health’ of the agricultural economy and at a macroeconomic level provides a picture of the economic stress on farming enterprises in recent years. Pearson's correlation coefficients and linear regression were used to investigate the relationship between terms of trade and farm suicide rates.
Results
Over the period 1988–97 there were 921 farm suicides identified in Australia. The farm manager group accounted for 621 (or 67.4%) of these suicides. Farm manager males represented 97.3% of these suicides. For the 300 agricultural labourer suicides, 96.7% were male. From 1990 onwards, male farm manager suicides have numbered over 50 per year. Overall, for this group the number of suicides ranged from 43 to 69. Male agricultural labourers are also represented, with frequencies ranging from 22 to 35 suicides for each year. Smaller frequencies in 1988 and 1989 for farm managers are evident, but not for agricultural labourers.
Approximately 48.5% of farm manager suicides occurred in ages over 55 years. In contrast agricultural labourer suicides occurred in the younger age groups, with the 15–39 years age group accounting for 61.4% of all agricultural labourer suicides. The relatively low occurrence of suicide in the female groups is also noted. As noted above female farm suicides will be underestimated using the current occupational classifications. Due to these small numbers, the 27 female suicides were not included in further analyses. Age-specific trends were also investigated based on broadly aggregated age groups (15–24, 25–34, 35–59, 60 +) for both farm managers and agricultural labourers. Most notable was an increasing rate for farm managers aged 60 + years across the 10-year period peaking at 106 per 100 000 in 1997. Similarly high rates and increasing trends were not evident in other age groups. There was, however, an increase in the 25–34 years age group for the years 1994 and 1995 peaking at 59 and 61 per 100 000, respectively, and a decreasing trend in the 35–59 years age group from 39 to 22 per 100 000. For agricultural labourers notable trends were not evident for those aged greater than 25 years. Rates were highest in the 15–24 years age group, peaking at 127 per 100 000 in 1989 and again at 76 per 100 000 in 1994.
Methods of farm suicide were also investigated based on the International Classification of Diseases (9th edition) external causes coding [37]. Overall, for male farm managers 80.8% of completed farm suicide could be accounted for by the methods of ‘hunting rifle’ (25.5%), ‘hanging’ (16.4%), ‘shot gun’ (13.4%), ‘other and unspecified firearm’ (13.2%), and ‘motor vehicle exhaust gas’ (12.3%). For male agricultural labourers these methods accounted for 81.0% of farm suicides, with ‘hanging’ representing a higher percentage than for farm managers (25.2%), and shot gun representing a lower percentage (9.7%). This is consistent with other reports of method [23]; however, suicides involving firearms are more frequent than suicides completed by hanging. Overall firearms accounted for 51% of male farm suicide, in comparison to 23% for the wider Australian male population for the same period [24].
Unadjusted rates per 100 000 for farm managers over the 10-year period ranged from 23.3 to 41.1 per 100 000. For agricultural labourers unadjusted rates ranged from 21.3 to 32.9 per 100 000. Age adjusted rates for farm managers range from 24.8 to 51.4 per 100 000. For agricultural labourers age adjusted rates ranged from 23.8 to 41.9 per 100 000. Comparisons with the national rate for the years 1988, 1990, 1992, 1994, 1996, and 1997 [10] reveal significantly higher rates, particularly for the farm manager classification. For farm managers, the rate ratio in comparison to the national rate increased from 1.53 to 2.19 over the 10-year period.
A relationship between male farm manager suicides and terms of trade is also evident—in general the farm manager rate increased from 24.8 to 51.4 per 100 000 while the terms of trade index for the same period declined from 100 to 82. This trend is not apparent for agricultural labourers and Australian males. Correlation and linear regression analyses revealed a strong negative correlation between male farm manager suicide rates and terms of trade (r = −0.75, p < 0.01) and an R2 of 0.51. In comparison there was no significant association apparent for agricultural labourers, nor was there any association with reported national rates.
Discussion
This study identified 921 farm suicides based on the agricultural classifications of farm managers and agricultural labourers for the period 1988–97. This is approximately 92 suicides per year, or one suicide every four days. Thus farm suicide occurs almost as frequently as work related death on farms – there is approximately one work related death on Australian farms every three days [25]. When work-related injury and death are considered together, the cost to the industry has been estimated to be 0.5–1.3 billion dollars per year [26]. In this context it is worth considering the potentially high economic cost of farm suicide.
Of the identified farm suicides the majority were male farm managers (67.4%). Most of these farm manager suicides were in older age groups, with 48.5% of suicides occurring in ages 55 years and older. Age specific rates over the period also suggest an increase in this older age group. For agricultural labourers, a much younger population, the trend was reversed with most suicides (61.4%) occurring to male agricultural labourers aged 15–39 years. Trends over the 10-year period in this group were less evident; however, rates were consistently higher in the 15–24 year age group with notable peaks in 1989 and 1994. The most common methods identified for both groups were firearms, particularly hunting rifles and shot guns, hanging and motor vehicle exhaust gas. These methods accounted for approximately 81% of all suicides in both farm manager and agricultural labourer groups. Firearms accounted for 51% of all male farm suicides, in comparison to 23% for the wider Australian male population for the same period [24]. The predominance of firearms as a method in the farming community is noteworthy, particularly when considered in the context of general rural suicide trends which indicate a decrease in the use of firearms as a method of suicide [27, 28].
Estimated direct age standardized rates per 100 000 for farm managers range from 24.8 to 51.4 per 100 000 for the 10-year period. For agricultural labourers, age standardized rates are slightly lower ranging from 23.5 to 41.9 per 100 000 for the 10-year period of interest. There are, however, current difficulties in establishing a definitive reference population for the farming population, and as such these rates are preliminary in nature. This method is likely to overestimate the farm suicide rate, given the larger male-female ratio present in the occupational population estimates in comparison to resident estimates; however these rates are lower than Australian farm suicide rates reported elsewhere [9]. The farm suicide rates in this paper are higher than those reported for the wider rural population, and are more similar to those reported for rural males aged 15–34 [27]. Analysis of the identified farm suicides in terms of the declining terms of trade for farmers revealed a significant negative correlation, with terms of trade accounting for 51% of the variance in a univariate linear regression analysis.
In terms of prevention, restricting access to means of suicide through population approaches, particularly where the method is prevalent, has been demonstrated to be successful [29, 30]. This preliminary analysis would suggest that firearms, especially rifles and shotguns, is an area of priority for this population. However for the farming population, this approach alone is unlikely to be effective given the accepted use of firearms within agriculture for controlling pests such as pigs, rabbits and foxes, and to put down sick or injured stock. Cumulative models of farm suicide prevention which address issues of access in terms of the interaction of a range of related antecedent factors relating to rural economy, family and social issues, alcohol and substance use, and service access and utilization will arguably engender more responsiveness from the farming population. Approaches which have focused on safe handling and storage of firearms have been received positively by rural communities, and have provided a context in which an awareness of issues pertaining to firearm suicide can be raised in collaboration with government agencies, health professionals and community groups [31].
Recent years have seen significant restructuring of the rural economy. The nature of Australian agriculture has been changing in terms of scale, diversity, and production processes, creating an environment that is less a lifestyle and more a commercially driven business enterprise. Understanding the social, psychological and health impact of such restructuring and the nature and causes of suicide requires an understanding of these fundamental changes to industry structure. These changes have resulted in differential exposures across farms and commodity groups, and are reflected in antecedent factors associated with work related injury (i.e. different production systems, different seasonal and market pressures). However the differential effect on work-related injury and mental health of such structural changes to the industry and the wider rural population are not known, and is an area of future research. Geographical correlates of agricultural production have shown that there are significant differences on key health indicators and suicide in the Australian rural population [32]. Australian farming differs significantly from other countries in many ways; however, the role of farm stressors identified in the United Kingdom [33, 34] and North America [35, 36] in Australian farm suicide needs further investigation. These stressors include financial problems, administrative and legislative pressures, production pressures and family problems.
Further research that is able to include suicide information for all farm residents will provide a clearer picture of suicide in the farming community. In many instances, farm family members and other farm residents who are equally exposed to the vagaries of the rural economy, and the stresses and strains of farm life are not identifiable in agricultural occupational classifications. Further research that investigates methods more specifically based on rates and comparison to national trends is also warranted. Contextual approaches to farm suicide research, and rural suicide research more generally, that take into account the inherent differences in work and social activity across various rural environments will provide an even better understanding of the nature and causes of farm suicide. Use of specific agricultural commodity price trends and other such external influences may be useful for further investigation in this context. The farming sector has been under continued pressure in recent years. Farm suicide and related mental health variables are increasingly becoming issues for concern in the community and industry. Research which begins to provide accurate information about the nature and causes of suicide in the farming community, and begins to unravel the antecedent factors associated with psychological distress, service access and service utilization will ensure that farm suicide is appropriately addressed.