Agribusiness Review - Vol. 11 - 2003

Paper 1
ISSN 1442-6951

Modelling the effects of a temporary loss of export markets in case of a foot and mouth disease outbreak in Australia
Preliminary results on costs to Australian beef producers and consumers

Liangyue Cao, Nico Klijn and Trish Gleeson
Australian Bureau of Agricultural and Resource Economics
GPO Box 1563, Canberra ACT 2601, Australia

• Introduction
• Loss of export markets and farm level impacts
• Zoning
• Model
• Model assumptions
• Cost of an FMD outbreak
• Simulation results
• No zoning (scenarios A to F)
• Zoning (scenarios G to H)
• Conclusion and discussion
• References
• Appendix 1: Mathematical specification of the model
• Appendix 2. Data and values of parameters

Abstract

The size of the cost to the Australian beef industry of a potential loss of major beef export markets following an outbreak of foot and mouth disease (FMD) in Australia is important in determining appropriate precautions. The size of this cost is evaluated with a dynamic bioeconomic model of Australian beef production, consumption and export trade. The model developed by ABARE represents forward-looking competitive behaviour of beef producers and traders on domestic and export markets based on perfect foresight. The model enables estimation of the cost due to a FMD outbreak under various scenarios regarding the duration of market closure and affected zones.

Introduction

In the meat industry access to many markets is restricted on the basis of whether or not products come from a disease-free country. Australia is relatively free of exotic diseases and, because of this, is able to export agricultural products to most other countries of the world. One of the most significant animal diseases is foot and mouth disease (FMD) (Donaldson 1987; Sanson 1994). It almost exclusively affects cloven-hoofed animals, both wild and domesticated. Cattle, sheep, goats and pigs are the species most commonly infected. Australia's current disease-free status would be undermined if its livestock industries, particularly its beef industry, became affected by FMD.

There are effectively two markets for meat in the world - markets that are free of FMD and markets in countries where the disease is endemic. The Pacific Basin market is made up of countries, which only accept imports from FMD free countries. These quarantine restrictions limit the supply of beef to the Pacific Basin and cause the returns on beef sales to the Pacific Basin markets to be higher than available for the same product in FMD endemic markets.

While FMD is a severe disease in its own right, for an exporting country like Australia the greatest economic impact would be due to the loss of export markets. Approximately 60 per cent of Australia's beef and veal production is currently sold on the world market and over 90 per cent of these are exported to countries in the Pacific Basin. Therefore, the short run impact of a disease incursion on the industry would be substantial, as alternative markets for a large quantity of beef would need to be found. High volumes of Australian beef redirected to these lower priced markets would depress prices further, feeding back to Australian farmers through lower prices at the saleyard.

Estimation of the cost to Australia due to an FMD outbreak has been the subject of a number of studies. Depending on assumptions used in the estimations and the coverage of costs defined, the costs estimated can vary significantly. Bunn (1993) estimated conservatively that an outbreak of FMD would cost Australia $10 million a day. ABARE recently estimated that the value of beef and veal exports would fall by $1209 million in the first year of an outbreak of FMD due to the loss of markets in the Pacific Basin and due to the need to seek buyers in other markets (ABARE 2001).

In the present study, a bioeconomic model is developed and used to quantitatively estimate the cost to the Australian beef industry of a loss of major FMD free markets in the event of an FMD outbreak in Australia. A broad outline of the model and results of model simulations under various scenarios are presented below.

Loss of export markets and farm level impacts

If an FMD outbreak were to occur in Australia, the most significant economic impact on the Australian beef industry would be due to the loss of major FMD free export markets. The impact on the Australian broadacre agriculture sector would be severe. Apart from the direct costs associated with eradication programs and compensation payments to producers, a significant decline in cattle prices would be expected as traditional markets for Australian beef and veal were closed off for a period of time.

Estimates of the costs of an FMD outbreak in Australia depend on a wide range of assumptions regarding the nature of the outbreak, methods and timing of containment and market access conditions. Previous research conducted by ABARE  (Lembit and Fisher 1992) contained estimates of the reduction in farm cash incomes due to an FMD outbreak, assuming producers maintained their normal longer term profit maximisation behaviour. Lembit and Fisher (1992) estimated that average broadacre farm incomes would fall by $25 700 per farm during the first year of an Australia wide FMD outbreak. In current dollar values this equates to a fall of $29 574 per farm. With almost 72 000 broadacre farms in Australia, broadacre farm income losses due to an FMD outbreak in Australia are estimated at around $2117 million for the first year.

ABARE has now developed a bioeconomic model of the Australian beef industry, which allows what if evaluations of a wide range of disease outbreak, containment and market access scenarios. The model focuses on estimation of costs due to the loss of major beef export markets.

Zoning

Zoning is a way of separating FMD affected from unaffected areas, where the unaffected areas are either not subjected to FMD related trade restrictions or subjected to such restrictions for a shorter period. The application of zoning for FMD in Australia was discussed at the 1994 National Workshop on Zoning for FMD (Anon 1994), where it was concluded that zoning could be used in the event of an FMD outbreak and that it would be particularly attractive where an outbreak takes some time to eradicate. Early studies (eg Barry et al. 1993) showed that the adoption of zoning could significantly reduce the economic impact of an outbreak compared to the situation without zoning. For example, Barry et al. (1993) showed that the adoption of zoning produced substantially smaller losses to the beef industry than those implied by the results of Lembit and Fisher (1992) from an analysis without zoning. 

The bioeconomic model developed by ABARE can be used to simulate various scenarios of durations of market closure and zoning. The model enables arbitrary sizes of the zones to be considered.

Model

Based on some assumptions about the relevant biological and economic conditions of the Australian cattle industry, a mathematical representation of the model is provided in Appendix 1

The model is a dynamic bioeconomic model, which represents forward-looking competitive behaviour of beef producers and traders on domestic and export markets based on perfect foresight. At each time step, the model estimates the age and sex composition of cattle herd, number of calves reared, slaughter rates, volumes of beef and prices consistent with a linear demand function. The model also contains estimates at each time step of feed costs, slaughter costs and total revenue as the product of beef produced and the price received.

The model uses a half-year time step. It considers 10 half-year age cohorts for non-breeding cattle (males and non breeding females) and 20 half-year age cohorts for breeding cattle (cows). These numbers of age cohorts are large, given that farmers generally keep cows up to around 7 years of age (14 half-years) and non breeding cattle up to around 3 years of age (6 half-years). The reason for using large numbers of age cohorts is to allow the model to have the flexibility to keep animals longer if a shock (such as FMD) occurs.

The model has an option to adopt zoning, that is, the model can be used to simulate both zoning and no zoning cases. In the case of zoning, two zones with arbitrary sizes are considered. A single domestic beef market is considered in the model in the case of no zoning, and two domestic markets in the case of zoning (one for each zone). There are two export markets in the model; one represents the FMD free market and the other the FMD endemic market. Each market has demands for two types of beef (premium beef and lower quality beef).

Below are some assumptions used for setting up data or parameter values that are used in subsequent model simulations.

Model assumptions

(1) Initial steady state equilibrium

In the absence of FMD, initial steady state equilibrium of the model is assumed to exist, and the system is assumed to stay in the initial steady state equilibrium. This equilibrium is characterised by all variables of the model being stable. 

At the prevailing prices and costs, no producer or trader wants to alter decisions about the number of female calves retained for breeding, the slaughtering age, or where to sell. The equilibrium depends on the assumed values of the model parameters (see Appendix 2 ). These parameters are saleable kilograms of beef per head by age and cattle type, feeding costs per head by age and cattle type, slaughtering costs (including other final costs) per head by cattle type, natural mortality rates, demand parameters on beef markets and the rate of discount. Growth in consumption and exports as a result of population and income growth is ignored in the analysis.

(2) Model calibration

The model is calibrated by allowing the parameters of feeding costs and slaughtering costs to be determined so that the initial equilibrium derived from the model is consistent with existing data on the volumes and prices of beef, while values of all other model parameters are given exogenously. Once these feeding and slaughtering costs are determined, they are then fixed as model parameters in all model simulations later on.

(3) Linear demand schedule

The demand schedule is assumed to be linear. The intercepts and slopes of the linear demand functions for two types of beef (premium beef and lower quality beef) at various markets are estimated based on the assumed elasticities and the observed initial data on volumes and prices of beef. The parameters of these linear demand functions are assumed not to change over time. 

(4) Live cattle

There are no exports of live cattle explicitly considered in the model developed. However, live cattle exports are approximated by their equivalent in exports of lower quality beef. During calibration of the model, the observed live cattle export values are converted to the volumes of lower quality beef by dividing by the observed prices of lower quality beef. The volumes of lower quality beef used in the model calibration are then the sum of the observed volumes of lower quality beef and the converted volumes of lower quality beef from the value of live cattle exports. 

Cost of an FMD outbreak

The cost of an FMD outbreak considered in this study only consists of the net cost to Australian beef industry and beef consumers, and does not include any costs related to control of FMD (eg any costs of the human resources involved in controlling the disease). The cost certainly depends on FMD outbreak scenarios, for example, how long the relevant markets are expected to close, whether zoning can be adopted, how many animals are affected, and so on.

Loss of major markets due to FMD, particularly the higher priced FMD-free export markets for some period of time, results in a temporary reallocation of beef produced to the remaining and much smaller export markets and the domestic markets. This will result in lower beef prices in the short term. In the longer term, a new steady state equilibrium, identical to the original steady state, will be established. The temporary reduction in prices and the expectation of prices reverting to initial steady state levels in the near future, results in adjustment in female retention and slaughtering decisions. The model is used to simulate the adjustment path that follows from the decisions by profit maximising producers and traders, operating on competitive markets, using their perfect foresight about market outcomes while discounting future benefits and costs.

The model estimates the net cost to Australia beef industry and beef consumers due to the FMD induced market closure as the net present value of losses and gains to the Australian producers and consumers over the period from the time of FMD outbreak to the time when the system returns to the original steady state.

Simulation results

A number of scenarios were simulated using the model which is implemented in the General Algebraic Modelling System (GAMS). The scenarios selected in this study (table 1) are used to test the FMD impacts on the Australian beef industry with different options which are zoning or no zoning, long closure periods or short closure periods of relevant markets, considering seasonality of beef production or not. In all scenarios, the simulation period is 60 half years (i.e. 30 years).

Table 1: Selected scenarios simulated in this study

Base scenarios with no FMD outbreak and the system staying in the steady state.
18 months (Jun): FMD outbreak in the January to June season; 18 months (Dec): FMD outbreak in the July to December season.
Victoria is assumed to be the FMD affected zone.

There are three base scenarios (scenarios A, D and G) reflecting different cases: scenario A considers no zoning and no seasonality, scenario D no zoning but seasonality, and scenario G both zoning and seasonality.

Scenarios B and C are chosen to compare the impacts of different durations of market closure periods, where 6 month longer closure period is used in scenario C than in scenario B. Scenarios E and F with an FMD outbreak in different seasons are used to test the impacts of consideration of seasonality in beef production. Scenario H is chosen to test the impact of zoning, for comparison with the no zoning scenario E. Detailed simulation results for each scenario are provided below.

No zoning (scenarios A to F)

If zoning is not adopted, then the FMD free export markets will be closed to all beef produced in Australia. Scenarios A to F in table 1 were simulated for this case.

(1) Scenarios A to C (no seasonality for varying closure periods)

In these scenarios it is assumed that there is no seasonal variation in beef production. To investigate the impact of the duration of the closure period, two scenarios were used. In scenario B there is a closure period of two half years (i.e. 12 months), while in scenario C there is a closure period of three half years (i.e. 18 months). Shown in figure 1 are the resulting time paths for the herds of breeding cows and non-breeding cattle over the 60 half years.

Figure 1: Time paths of herds with market closure periods of two half years (scenario B) and three half years (scenario C), compared with the steady state time path (scenario A)

The figures show that immediately after the market closure, the herd of non breeding cattle rises to reflect the accumulation of non breeding animals due to the closure of major premium beef markets; while the herd of breeding cows declines (slaughter rates increase) during the FMD outbreak because it is not economic to keep cows longer due to higher feeding costs per head. It should be mentioned at this stage that the model uses an assumption that there is no distinction between premium beef and lower quality beef during the market closure period.

This implies that the higher supply from the increased slaughter of cows is supplemented by the (lower) supply from the reduced slaughter of non-breeding cattle on the remaining beef markets during the closure period.  From figure 1, the longer closure period (scenario C) results in a larger accumulation of non-breeding cattle, and more slaughtering of breeding cows. This time-path for herds is consistent with the time-path of prices shown in figure 2.

Figure 2: Time paths of prices with market closure periods of two half years (scenario B) and three half years (scenario C), compared with the steady state time path (scenario A)

The closure of major export markets reduces the price for both premium beef and lower quality beef. The longer closure period results in even lower prices in the early years of the FMD outbreak. Both figure 1 and figure 2 show that the system gradually returns to the initial steady state, after the shock due to FMD.

The net economic cost to the Australian beef industry and consumers for the scenarios is reported in tables 2 and 3, where the cost is estimated to be around $2408 million in present value terms over 30 years in scenario B. The cost increases to $3792 million in scenario C with a 6 month longer closure period than that in scenario B. The cost in the first year of the FMD outbreak is around $2703 million in scenario B, while this cost slightly decreases to $2695 million in scenario C. This decline in costs is partly explained by increased consumer surplus generated by lower domestic beef prices for the extended closure period.  This does not contradict that the total cost over 30 years in scenario C is higher than that in scenario B, because in the subsequent years the costs in scenario C are higher than those in scenario B due to the longer closure period.

(2) Scenarios D to F (seasonality)

In reality beef production follows a seasonal pattern with lower production in the January to June half year. The lower production in the January to June season may be due to the holiday season of Christmas and New Year. In these scenarios, seasonality is added to the model, which allows the steady state data to vary between the January-June season and the July-December season. In this case, the equilibrium is a two-period cycle from the state in January-June to that in July-December, and then back to the same state in January-June, and so on. The length of closure period, 18 months, is the same as in scenario C.

The time-paths of herds and prices for the scenarios D and E with seasonality compared with the results for scenario C without seasonality are shown in figures 3 and 4, where it is assumed that FMD occurs in the January-June season. The results show little difference between the case with seasonality and the case without seasonality.

Figure 3: Time paths of herds for an 18 month closure period, both with and without seasonality (ie scenario E and scenario C), compared with the steady state time path with seasonality (ie scenario D)

Figure 4: Time paths of prices for an 18 month closure period, both with and without seasonality (ie scenario E and scenario C), compared with the steady state time path with seasonality (ie scenario D)

The small difference between the results with and without seasonality implies that the timing of an FMD outbreak, that is, in January-June or July-December, will not have much effect on the economic cost to the beef industry and consumers. To see this, a simulation assuming an FMD outbreak in July-December (scenario F) was also carried out. Figures 5 and 6 show the comparison of results of an FMD outbreak in January-June (scenario E) and the results of an outbreak in July-December (scenario F).

Figure 5: Time paths of herds with an FMD outbreak in the January-June season (scenario E) and in the July-December season (scenario F), compared with the steady state time path.

Figure 6: Time paths of prices with an FMD outbreak in the January-June season (scenario E) and in the July-December season (scenario F), compared with the steady state time path.

Again, for all scenarios after the shock due to FMD the system gradually returns to the initial steady state. The net economic cost to the Australian beef industry and beef consumers for these scenarios is reported in tables 2 and 3, where the cost is estimated to be around $3900 million in present value terms over 30 years in the case of an FMD outbreak in the January to June season (scenario E), and around $3846 million in the case of an FMD outbreak in the July to December season (scenario F). As mentioned earlier, the results show little difference in the costs of an FMD outbreak in the January to June season and in the July to December season. This implies that the timing of FMD outbreak does not have much impact on the estimated economic cost of an outbreak although there is some variation in beef production between seasons. The cost in the first year is around $2698 million in the case of an FMD outbreak in the January to June season, and around $2719 million in the case of an FMD outbreak in the July to December season. Comparing the costs for scenarios E and F with seasonality with the cost for scenario C without seasonality, the costs for scenarios E and F are higher than for scenario C. 

Zoning (scenarios G to H)

Scenarios were constructed to analyse the effect of zoning on the economic costs of an FMD outbreak. It is assumed that there are two zones, and for illustrative purposes, one zone is chosen to be Victoria and the other is the rest of Australia. Seasonality is included in these scenarios. To make the results comparable with the results from the scenarios without zoning, the closure periods are set at three half years for the FMD affected zone (which is the same as in the corresponding scenario without zoning (i.e. scenario E) and a half year for the FMD free zone. It is assumed that an FMD outbreak occurs in the January to June season. Results for the FMD affected zone (i.e. Victoria) are shown in figures 7 and 8, and results for the FMD free zone (i.e. the Australia excluding Victoria) are shown in figures 9 and 10.

Figure 7: Time paths of herds in the FMD affected zone (ie Victoria) with an FMD outbreak in the January-June season (scenario H), compared with the steady state time path (scenario G).

Figure 8: Time paths of prices in the FMD affected zone (ie Victoria) with an FMD outbreak in the January-June season (scenario H), compared with the steady state time path (scenario G).

Figure 9: Time paths of herds in the FMD unaffected zone (ie Australia excluding Victoria) with an FMD outbreak in Victoria (scenario H), compared with the steady state time path (scenario G).

Figure 10: Time paths of prices in the FMD unaffected zone (ie Australia excluding Victoria) with an FMD outbreak in Victoria (scenario H), compared with the steady state time path (scenario G).

Overall the dynamic behaviour with zoning is similar to that shown in the scenarios where zoning is not considered. The impact in terms of changes in herds on the FMD affected zone is greater than on the FMD free zone. The greater impact on the FMD affected zone is caused by the longer period of market closure. To compare the economic impacts for the case with zoning and the case without zoning, see tables 2 and 3 for detailed estimates of economic costs.

For the FMD affected zone there is a loss of around $448 million to beef producers in present value terms over 30 years, however, this loss is offset by a gain of around $485 million to beef consumers due to the low beef prices in earlier years following the FMD outbreak. So, there is a small net gain of $37 million to beef producers and consumers in present value terms over 30 years. For the FMD free zone the net cost is around $1582 million to beef producers and consumers in present value terms over 30 years, where there is a loss of around $2449 million to beef producers and a gain of around $867 million to beef consumers. Also, note that the FMD free zone (i.e. Australia excluding Victoria) is much bigger than the FMD affected zone (i.e. Victoria).

For the whole of Australia, the net loss is around $1545 million in present value terms over 30 years with the adoption of zoning. This loss is significantly smaller than the loss of $3900 million for the corresponding scenario without zoning (i.e. scenario E). Looking at the costs in the first year only of an FMD outbreak, the cost for the FMD affected zone is around $433 million, while for the FMD unaffected zone the cost is around $1228 million. Taking Australia as a whole, the cost is around $1662 million in the first year of an FMD outbreak. Again this cost is significantly smaller than the cost of $2698 million for the corresponding scenario without zoning (i.e. scenario E). These results show that the application of zoning can have a significant impact on reducing the costs to Australia if an FMD outbreak were to occur.

Table 2: Present values (million Australian dollars) over 30 years

'cf A' means compared with Scenario A. Similar meaning for 'cf D' and others.

Table 3: Present values (million Australian dollars) in the first year of FMD outbreak

'cf A' means compared with Scenario A. Similar meaning for 'cf D' and others.

Conclusion and discussion

A dynamic bioeconomic model was developed to estimate the net economic costs to the Australian beef industry and beef consumers were an FMD outbreak to occur in Australia. A number of scenarios were tested to demonstrate the model capabilities. Zoning and seasonality were analysed under different outbreak conditions.

The results showed that a longer closure period of FMD free export markets would result in significantly higher net costs to the Australian beef industry and beef consumers. Adding seasonality into the model to more accurately model Australian beef production results in a slight increase in the estimated loss from an FMD outbreak. With respect to zoning, the model results showed that the adoption of zoning could significantly reduce the net costs to the Australian beef industry and beef consumers were an FMD outbreak to occur.

Two issues need to be addressed. One is destruction of FMD affected animals, and the other is the value of the price elasticities of demand. In the current model, there is no consideration of destruction of FMD affected animals. However, this issue is to be considered in further modelling work, given that destruction of affected animals was used in the UK FMD outbreak to limit the spread of the disease. The price elasticities of demand are assumed to be -1 for all domestic and export markets in the present study. Other values of the price elasticity of demand for beef in FMD endemic markets will be considered in a further study, after more investigation on the demand and supply conditions in these markets.

One other point should be mentioned. The model assumed competitive behaviour of beef producers and traders on domestic and export markets. However, in principle, the model can be modified to deal with some forms of non-competitive behaviour of producers and traders.

References

ABARE 2001, FMD: short run macroeconomic and intersectoral impacts on the Australian economy of a foot and mouth disease outbreak, Report to the AFFA Foot and Mouth Disease Task Force, September.

Anon 1994, Government/Industry Workshop on Zoning for Foot-and-Mouth Disease, Melbourne 3-6 October.

Barry, G., Shaw, I., Beare, S. and Short, C. 1993, The costs and consequences of an FMD outbreak: Implication of zoning policies for Australian broadacre agriculture, Paper presented at the Australian Veterinary Association Conference, Gold Coast, 16-21 May.

Bunn, C. 1993, 'Why could the introduction of an exotic disease into Australia be a disaster?' The Macedon Digest , vol.8, pp. 1-5.

Donaldson, A.I. 1987, 'Foot-and-mouth disease: the principal features', Irish Veterinary Journal , vol. 41, pp. 325-7. 

Lembit, M.J. and Fisher, B.S. 1992, 'The economic implications of an outbreak of foot and mouth disease for Australian broadacre agriculture', In M. J. Nunn and P. M. Thornber (editors), Proceedings of the national symposium on foot and mouth disease , Canberra, 8-10 September, pp. 83-90.

Sanson, R.L. 1994, 'The epidemiology of foot-and-mouth disease', New Zealand Veterinary Journal , vol. 42, pp. 41-53.