Cost–benefit analysis of a disease control programme with special reference to ticks and tick-borne diseases in the former Venda region

Abstract

This study is based on a cross-sectional survey of 125 small-scale cattle farmers interviewed in the Venda region of the Limpopo Province of South Africa. It revealed a 3 per cent mortality rate in spite of the existence of a dipping programme. Cost–benefit analysis revealed a cost–benefit ratio of 0.8 (i.e. <1) indicating that the control of ticks and tick-borne diseases by the government is not economically justified. However, because of the broader socio-economic benefits it provides, the dipping of cattle still deserves governmental support. In addition, the provision of tick control services by the government leads to a socially optimal level of supply of animal health services in general. Sensitivity analysis gives a cost–benefit ratio of 1.2 when it is assumed that the mortality rate would have been 10 per cent without the control programme.

1. Introduction

Profound changes are under way in delivery systems for animal health inputs and services in much of Africa. Livestock health service delivery in many developing countries is undergoing privatisation as part of an international restructuring programme for economic development. Of relevance to South Africa's developing areas is the recent government withdrawal of dipping subsidies to small-scale cattle farmers. This has had some negative impacts on the farmers and these are significant if the multiple effects are also considered.

The objective of this paper is to demonstrate the costs and benefits of tick control at the farm level. The study was conducted within a cost–benefit analysis framework, and the paper begins with an outline of this methodology. The methodological discussion is followed by a cursory discussion of various ticks and tick-borne disease control measures applied in South Africa's developing areas. The ‘with’ and ‘without’ dipping scenarios are incorporated into the analysis to provide a basis for cost–benefit analysis. The study proceeds with a comparative analysis of costs and benefits for each strategy to determine whether it is economically justified to control ticks and tick-borne diseases. The role of the government in the provision of animal health is highlighted and the study concludes with some recommendations.

2. Methodology

On the advice of the former Venda Department of Agriculture (Veterinary Service Division) in the present Venda region of Limpopo Province, it was decided that surveys should be conducted in all the veterinary zones, namely, Red Line, Yellow Line and the Open area. However, for socio-political reasons surveys could not be conducted in the Redline area. Two-stage sampling was performed in this study. Firstly, three areas, each of which contains a diptank for tick control purposes, were chosen within each zone. Secondly, cattle within each diptank area were sorted into four categories by number of cattle: 1–10, 11–20, 21–30, and more than 30 head of cattle. Then, within each category, simple random sampling was performed using random number tables. The aim was to obtain 25 respondents from each diptank area using the following formula:

• n = number of cattle owners within each stratum

• N = total number of cattle owners in a diptank area

The actual number of farmers who participated in the survey was 125, which was less than the targeted number of 150. However, as the sample number was about 80 per cent of the target it was considered large enough to be representative.

Various techniques are available for assessing the impact of agricultural research and developmental projects and they are well documented in Anandajayasekeram et al. (1996). They include modified peer review, user surveys, the cost–benefit method, patent analysis, mathematical programming and econometric methods. The study used a cost–benefit analysis approach because this technique is much more appropriate for assessing past projects than for ongoing or future research. Its main flaw is that because it bases its assessment on quantification in monetary terms it does not take into account non-monetary benefits that farmers derive from cattle production; for example, some small-scale cattle farmers keep cattle only for prestige.

3. The control of ticks and tick-borne diseases

Control measures against ticks and tick-borne diseases were applied on a large scale in southern Africa following the spread of East Coast Fever (ECF) from East Africa (Norval, 1994). After the eradication of ECF essentially two options were open to southern African countries: they could either control the remaining major tick-borne diseases (babesiosis, anaplasmosis and heartwater) by vaccination and move towards reduced tick control and endemic stability, or they could continue to control the diseases by intensive dipping. It was established that the most efficient and economical way to control ticks was to target treatment at the parasitic stage and short interval dipping of livestock became the standard method of tick control.

In South Africa, compulsory dipping was abolished and the choice was left to individual farmers (Norval, 1994). However, in some of the previously autonomous South African homelands compulsory dipping carried out by the government on communal grazing continued to be enforced through legislation. This was the case in Venda. It was stipulated in the Venda Government Gazette that: ‘By law a stockowner is now compelled to produce his cattle for inspection and tally recording by a stock inspector at the stock inspection point (diptanks) at least every 14 days’ (Republic of Venda, 1984), which implied that the cattle must also be dipped, although this was not directly stated. Cattle were brought to the diptank for the two purposes mentioned (inspection and tally), but also to be immersed, at considerable expense relative to the farmers' contribution. (Acaricides are provided at a highly subsidised rate.) According to Lawrence (1996), there is no justification in insisting that every animal be produced as often as every week and immersed in a tank at considerable expense for the purpose of controlling other dangerous diseases. Compulsory dipping may at one time have been thought to be leading towards eventual tick eradication, which would have provided a possible case for public funding, but today there is a growing perception that eradication is an unlikely prospect.

Cattle were dipped either on a weekly or fortnightly cycle depending on the tick challenge of a particular area, although Bachmann (1992) argues that less than 10 per cent of the cattle in Africa are dipped with any regularity. In Venda, cattle were supposed to be dipped depending on tick challenge, but this is complicated in a communal dipping set-up because farmers only go to the diptank for the control of ticks and tick-borne disease and do not seem to care about the control of other diseases. In general, dipping frequency varies by region depending on, for example, water and acaricides availability, and usually when all the necessary dipping inputs are available farmers do dip their cattle whenever they present them, whether weekly or fortnightly. The farmers see the dipping services as being for the control of ticks and tick-borne diseases only, and as soon as you start providing other services such as inspection they are reluctant to present their cattle to the diptank. The intention of the dipping policy is to dip them often, but lack of other dipping inputs such as water, etc. means that this does not happen. To the farmers, the results of controlling other diseases through the use of diptanks are not discernible and the only incentives to make them present their cattle for the control of these other diseases is simply to dip their cattle.

There is a large variety of acaricides (‘tickicides’) to choose from, each having its own application and management system (Dipping Policy Survey, 1997). The ones commonly used are Grenade, Triatix and Clout (pour-on). Pour-on is generally used in cases where there is a shortage of dipping water and a heavy tick infestation, and when cattle are not in a good enough condition to be capable of swimming. Unlike Grenade and Triatix, which are provided by the government, Clout is donated by Hoechst as part of the Reconstruction and Development Programme package. It is a worrying fact that resistance of ticks to the available acaricides is on the increase (Dipping Policy Survey, 1997).

The main reason for the construction of diptanks was the outbreak of ECF, whose control was dependent on the control of ticks. Once ECF was brought under control, it was apparent that dipping was beneficial to the farmer in other ways, as well as to the veterinary services. The gathering of animals at dipping venues offers opportunities to veterinary services to perform the following important functions:

• Disease surveillance. The physical concentration of cattle at diptanks is used by the state to perform compulsory inspection for controlled diseases, such as foot-and-mouth. Surveillance of foot-and-mouth disease is a function of international, national, provincial and local importance and is based on compulsory inspection at various intervals and movement control (through quarantines) of various intensities in the whole of the foot-and-mouth disease controlled area. The communal cattle dipping system makes it possible to check animals for condition, general health, production and reproduction, and monitoring of mortality. It is hardly possible, in fact, to perform diagnostic testing of certain controlled diseases, such as brucellosis, without the communal cattle-dipping set-up. However, the danger of communal diptanks is that they can act as centres for disease transmission.

• Extension. Dipping at communal diptanks is also seen as an opportunity for effective extension, education, training, practical demonstration and the collection of census data. Similarly, dipping attendance, inter alia, is an important forum for interface between cattle owners and government veterinary officials where cattle statistics are collected and movement permits authorised. Under this system the farmer is obliged to explain the whereabouts of absent cattle. Without this opportunity, the costs of performing extension as effectively would be enormous, making it unaffordable and unachievable.

• Extra benefits. Dipping venues also offer some unintended services. It has been observed, for example, that nowadays they also serve as a marketplace for private buyers. These buyers, for example butchers, use diptanks as a place to obtain animals for slaughter, to negotiate terms of trade and bargain to reach a mutually satisfactory price with the cattle owners. It can be argued that in the absence of reliable formal markets dipping venues provide a useful marketplace for both buyers and sellers, with the advantage of reducing the transaction costs to both.

4. Costs and benefits of alternative control programmes

The outbreak of ECF in the early nineteenth century presented the Venda Government (then the South African Government) with only two basic options: to control or not to control. The assessment of an animal health project takes into account the incidence of disease outbreaks before and after the project, and the losses which would have resulted from failure to implement the project (Sidibe, 1981). In this section the two options/scenarios are examined and are referred to as:

• A ‘do nothing’ strategy (‘without’ dipping scenario)

• A control strategy (‘with’ dipping scenario)

Emphasis is placed on comparing the costs and benefits of each strategy at the household level.

4.1 A ‘do nothing’ strategy

The introduction of exotic breeds made many of the cattle of Africa vulnerable to tick-borne diseases. Because of this breed susceptibility it would probably have been irrational at the time to adopt this strategy for economic reasons. Had this strategy been adopted, tick-borne diseases would have occurred in epidemic proportions, but as time progressed cattle would have adapted and the situation of enzootic stability would have developed (personal communication, AM Spickett, Protozoology Division, Onderstepoort Veterinary Institute, Onderstepoort, Pretoria, 1998). The existence of enzootic stability implies that control could have been selective, strategic and focused only on susceptible target cattle populations.

A ‘do nothing’ strategy would have incurred no additional government expenditure and farmers would have retained the use of their cattle, although the cattle would not have performed as efficiently as before the introduction of the disease, and more income would have been forgone because of reduction in cattle products than would be the case with the control strategy discussed below. This is because of the high morbidity and mortality rate, estimated to be 4 per cent. Although this strategy was not adopted either in Venda or in the rest of South Africa, an effort should be made to estimate the losses that would have occurred in the absence of control and then estimate the extent to which these losses would be reduced by the strategy followed.

4.2 A control strategy

One of the most significant consequences of the ECF control programmes was the introduction of compulsory dipping in acaricides throughout the infected areas. Whether eradication of ECF can be attributed entirely to the dipping programme and the other control measures that were applied, or whether climatic changes brought this about, or at least facilitated the elimination of a specific population of ticks responsible for transmission of ECF, is open to speculation (Norval, 1994).

The benefits of this strategy are higher than those of the ‘do nothing’ strategy. This is because of a relatively low morbidity and mortality rate of 3 per cent in the sampled region. Most developing regions were left with a tradition of short interval dipping. If properly applied, dipping provides very good control of all tick-borne diseases.

5. Derivation of costs and benefits estimates

5.1 The ‘without’ dipping scenario

One of the most important components of a cost–benefit analysis of the diseases control programmes is the estimation of the number of animals lost through disease. Estimating the number of individual animals lost is always a problem (Beal, 1981), and the problem is worse in a situation without a control programme – i.e. a ‘without’ dipping programme, for comparison – because dipping has been in operation comprehensively since the start of the century. Moreover, it is difficult, if not impossible, to obtain access to data from the period before the dipping programme. This lack of data resulted in one major difficulty: how to determine accurately the effects to be assessed, the costs and benefits, and the exact method to be used to measure these effects. It was decided to assess only direct benefits that appear at the farm level in terms of milk and meat production gains/losses, low productivity due to longer calving intervals, and impacts on draught power.

The derivation of benefits for the ‘without’ scenario was performed on the basis of various estimates by using the Delphi technique (referral). Because experimental data comparing the effect of ticks and tick-borne diseases on livestock productivity between tick-free and tick-infested herds is unavailable for South Africa, estimates of the impact on productivity done by the Food and Agriculture Organisation in Zambia were used with minor adjustments (Pegram et al., 1993).

Table 1 presents various figures used to estimate the costs and benefits for both milk and meat (beef). Based on the income from milk production as reported in Table 1, the impact on morbidity and mortality was calculated for the ‘without’ situation as follows (i.e. income gained from milk production):

It was estimated that on average the mortality rate would have stabilised at 4 per cent without the control programme (personal communication, AM Spickett, Protozoology Division, Onderstepoort Veterinary Institute, Onderstepoort, Pretoria, 1998). However, the initial mortality rate would have been 10 per cent without dipping at the beginning of the century. This possible decline in mortality is based on the assumption that breeds with a strong immune system (predominantly the Nguni breed) would have survived better and on the basis of ‘the survival of the fittest’ would have dominated the national herd.

Table 1: Values for the possible calculation of milk and meat benefits

Dipping scenarios	Average live- body weight (Kg)	Beef price/ kilogram* (R)	Average herd size	Cows per herd composition (%)	Mortality rate (%)	Value of milk/year (R)	Value of a cow/year (R)
‘With’	241	5	329	39	3	526	1152
‘Without’	234	5	326	38	4	405	1036
*Source: Author's survey data.

Income gained from beef production is calculated as follows:

For a small-scale traditional farmer the selling price of a live cow is not a function of weight. However, it is assumed that price is a function of weight in this analysis so that the price of beef per kilogram can be multiplied by the total weight of an animal. Moreover, the traditional farmer sometimes salvages a sick animal for meat. In this analysis all reported deaths are considered to be unsalvaged animals.

With regard to draught power, two significant assumptions had to be made to perform the analysis, namely:

• Only oxen are used for ploughing.

• Infected oxen are entirely incapable of ploughing.

Oxen were grouped into two breed categories, indigenous and exotic, so as to attach the infection probability for each group. Grouping of breeds was done using the sampled breed percentage composition. Nguni constitute 65 per cent of the sampled herd and it was assumed that Nguni is the only indigenous breed. Again it was assumed that both breeds have an equal infection probability of 25 per cent without dipping. The value of oxen indicated in Table 2 is for an average herd of 18 oxen after taking into consideration the infection probability. The average number of oxen for the sampled herd is 24.

Table 2: Comparison of costs and benefits for the ‘with’ and ‘without’ dipping scenarios (1997)*

‘With’ dipping scenario
Costs (R)		Income (R)
Government costs		Milk: sold	67 854
Motor transport allowance, maintenance and insurance	11 538	Mortality	−2 025
			65 829
Heavy machinery (water provision)	661	Meat: sold	69 890
Stationery	236	Mortality	−11 893
Livestock maintenance (acaricides)	15 535		57 997
Diptank maintenance	21 000	Draught power	6216
Total	48 970
Farmers' costs
Acaricides	1 164
Purchased drugs	682
Total	1 846
Salary	6 341
Total Cost	57 157	Total income	130 042
		Net income	72 885
‘Without’ dipping scenario
Costs		Income
		Milk: sold	50 171
		Mortality	−2 007
			48 165
			66 690
		Meat: sold	−15 257
		Mortality	51 433
		Draught power	4662
		Total income	104 260
		Law calving productivity	−7 900
		Total income	96 360
		Cost–benefit ratio	0.8
*This information was derived from the government veterinary budget.

Ticks may be responsible for loss in udder quarters and may increase the calving interval. The latter is the most significant parameter in determining herd productivity and hence profitability. This type of loss is better estimated on a survey basis (Pegram et al., 1993). Low productivity in milk production due to a longer calving interval for the infected cow is estimated as follows (i.e. income forgone because of the longer calving interval):

5.2 The ‘with’ dipping scenario

The derivation of benefits for this scenario is similar to the ‘without’ scenario. Unlike the ‘without’ scenario, this relies on the survey results. Survey results reveal the mortality rate to be 3 per cent. Mortality rates exhibit a strong seasonal pattern, being highest in the rainy season when ticks are most active. No attempt was made here to differentiate the various mortality rates for the different seasons of the year. To measure the mortality rate, sampled farmers were asked about the number of cattle deaths caused by tick-borne diseases. It is acknowledged that the reliance on farmer diagnosis has a potential for error depending on, inter alia, familiarity with the disease, memory, and the desire to attract veterinary attention. In communal areas mortalities are not always reported or, if reported, seldom investigated. Moreover, disease reporting in general in South Africa is unsatisfactory, which results in underreporting of most diseases, and unreliable statistics (National Department of Agriculture, 1996/97). Although these limitations render the estimations less reliable and accurate, they serve as the only basis on which the analysis can be built. Both milk and meat production incomes can be obtained using the values in Table 1.

For a small-scale traditional farmer the costs of tick control can be divided into two categories:

• Those that the government incurs.

• Those incurred by stockowners.

Data on government expenditure for ticks and tick-borne diseases control was obtained from the government veterinary budget, reports and records in the Department of Veterinary Services. Costs incurred by the government include vehicle maintenance, insurance and allowances, stationery, dip maintenance and heavy machinery for water provision. The cost of water provision is low because the location of many diptanks was determined by the available water sources. The costs to individual farmers of control of ticks and tick-borne diseases are fairly straightforward and are based upon expenses for animal treatment, largely for purchased drugs, and for application of acaricides. These costs were obtained from sampled farmers' interviews and are shown in Table 2. Treatment of the disease is regarded as an ex post use of resources to restore animal performance to the level it was before the disease occurred. Whether it is worth doing, as opposed to accepting the reduced productivity or even culling the animal, is an economic decision.

6. A comparison of the ‘with’ and ‘without’ scenarios

Project impact analysis attempts to value costs and benefits that arise and compare them with the situation as it would have been without the project (Gittinger, 1982). Both costs and benefits are clearly shown in Table 2, with milk and beef calculated in terms of mortality and a decrease in the productivity of the herd (i.e. morbidity). The income forgone for the ‘without’ scenario owing to mortality is higher than the income forgone for the ‘with’ scenario. The difference between the ‘with’ dipping scenario net income and the net income balance for the ‘without’ scenario gives an incremental net loss of R23 475 per sampled herd. This amount represents the loss in income incurred in spite of the existence of the dipping programme. Disease control is economically justified only if the estimated benefits outweigh the costs incurred, i.e. the cost–benefit ratio must be greater than or equal to one. Table 1 shows a cost–benefit ratio that is almost equal to one (0.8). This ratio was calculated as:

According to Morris and Meek (1980: 165), ‘monetary values must be used with the caution that the numerical values obtained in an economic analysis should be seen principally as a basis for ranking strategies, not as representing the actual benefit which will be achieved under all circumstances’.

This ratio suggests that the control of ticks and tick-borne diseases by the government is not economically justified and probably needs to be a private sector responsibility. However, in establishing the appropriate roles for the public and private sector in the livestock services industry, it is necessary to obtain a clear understanding of the nature of the tick control service. The economic nature of the service will determine not only whether private delivery will be feasible but also whether private provision will result in a socially optimal level of supply (Kirsten & Randela, 1998).

The control of ticks is a private good with externality. This implies that the control of ticks by an individual farmer also benefits other farmers by reducing the tick population and the chances of other farmers' cattle getting tick-borne diseases. Because tick control service delivery thus benefits ‘free riders’, there will be a tendency towards under-provision or no provision of this service when the production decision is profit motivated. Thus, private firms will have no incentive to provide this service because it will not be in the interest of any individual to pay for it.

Furthermore, although the cost–benefit ratio is slightly less than one, this ratio can be improved by reducing the costs involved. This can be done by moving from intensive dipping to a strategic acaricides application. Such a cost decrease might result in one of two effects: either the same output can be produced more cheaply or the savings can be converted into increases in output. The strategic control of ticks is based on the fact that ticks exhibit a seasonal cycle and that concentration of acaricides application during the peak month of tick activity will effectively interrupt the tick feeding cycle and reduce the tick population, thereby reducing the number of engorged females which lay eggs to perpetuate the cycle.

This dipping regime is possible if it is taken into consideration that livestock owners are more particular about regular immersion of their cattle during the spring/summer period (September/October to March/April), when the presence of ticks and the damage they cause are obvious, than in winter when only few or no ticks are seen. The presence of ticks on cattle in winter varies from region to region, from few to none, depending on the degree of dryness and severity of winter cold. Thus in some areas there may in fact be no need to dip cattle in winter. It appears, therefore, that if the strategic dipping regime was adopted there would be a significant decrease in dipping costs and an improvement of the cost–benefit ratio. The magnitude of a decrease in dipping costs depends on the knowledge of the optimal tick control. In addition, the range and magnitude of physical losses avoided, inter alia, depends on the control system and technique and the success with which it is implemented.

7. Sensitivity analysis

The mortality rate for the ‘without’ dipping scenario is, however, subject to debate and is based on optimistic assumptions. Sensitivity analysis is therefore conducted to determine the robustness of cost–benefit ratio against possible changes in the mortality rates. The sensitivity analysis presented in this section addresses the effects of changes in the mortality rate to 10 per cent and the results are shown in Table 3. With a 10 per cent mortality rate without the control programme the cost–benefit ratio increases from 0.8 to 1.2. Thus, it is evident that the cost–benefit ratio is sensitive to changes in the mortality rate. The 10 per cent mortality rate was chosen assuming that it would probably have been the second lowest mortality rate to prevail in the absence of the control programme.

Table 3: Sensitivity analysis of a change in the mortality rate

‘With’ dipping scenario
Costs (R)		Income (R)
Government costs*		Milk: sold	67 854
Motor transport allowance, maintenance and insurance	11 538	Mortality	−2 025
			65 829
Heavy machinery (water provision)	661	Meat: sold	69 890
Stationery	236	Mortality	−11 893
Livestock maintenance (acaricides)	15 535		57 997
Diptank maintenance	21 000	Draught power	6 216
Total	48 970
Farmers' costs	1 164
Acaricides	682
Purchased drugs	1 846
Salary	6 341
Total
Total Cost	57 157	Total income	130 042
		Net income	72 885
‘Without’ dipping scenario
Costs		Income
		Milk: sold	38 418
		Mortality	−3 842
			34 576
		Meat: sold	66 690
		Mortality	−35 802
			30 888
		Draught power	4662
		Total income	70 126
		Law calving productivity	−7 900
		Total income	62 226
		Cost–benefit ratio	1.2
*This information was derived from the government veterinary budget.

The sensitivity analysis also shows an incremental net benefit of R10 658 per sampled herd. It is important to note that the incremental net benefit derived represents the loss in income that would have resulted from failure to implement the dipping programme.

A sensitivity analysis based on various mortality rates, i.e. worst-case scenario, was also performed. In addition to the 4 and 10 per cent mortality rates already analysed, Table 4 further shows the probability of occurrence for various mortality rates together with their respective cost–benefit ratios. It is evident from Table 4 that in the ‘without’ dipping scenario the would-be mortality rate is inversely related to the probability of occurrence. This is because a higher mortality rate, e.g. 21 per cent, is relatively a worst-case scenario that is unlikely to occur, hence the low probability of 2 per cent. At 15 per cent and 21 per cent mortality rates the cost–benefit ratios are 1.5 and 2 respectively.

Table 4: Impact of varying mortality rates on the cost–benefit ratio for the ‘without’ dipping scenario

Mortality rate (%)*	Probability of occurrence (%)*	Cost–benefit ratio
4	80	0.8
10	40	1.2
15	10	1.5
21	2	2
*Source: Spickett (unpublished data, 1998).

Attempts to supplement this analysis with other measures of project value, internal rate of return and net present value were made. But lack of both benefit and cost series of data as well as data on initial investment costs precluded the performance of the analysis. This lack of data surely underscores the importance of institutionalising project monitoring and evaluation within the South African agricultural research system.

However, wherever programmes involve public expenditure the authorities responsible now require, inter alia, clear demonstration that programmes will show a net benefit and the extent of such benefits needs to be indicated. In performing analyses of ongoing programmes, there has to be a comparison of projected future progress with what would happen in the future if the programme was discontinued (Beal, 1981). Therefore, what is of relevance today is the analysis of what will happen in future if dipping can be completely discontinued. It is estimated that in the first, second and third year farmers are likely to face an average mortality rate of 10, 6, and 4 per cent respectively (personal communication, AM Spickett, Protozoology Division, Onderstepoort Veterinary Institute, Onderstepoort, Pretoria, 1998). Thus, it will take only three years for the herd to develop an acceptable immune system to the extent that the mortality rate can stabilise at 4 per cent.

The reason for the high mortality rate in the first year is that dipping to some extent destroyed the immune system of especially the indigenous African breeds. In addition, it resulted in an interruption in the transmission of tick-borne pathogens and led to the establishment of a susceptible cattle population followed by the loss of the existing situation of enzootic stability. Considering a 10 per cent mortality rate, of the total sampled herd of cattle (1976), 198 cattle are expected to die of tick-borne diseases, resulting in a R228 096 (198 × R1 152, see Table 1) loss to the sampled cattle owners. Obviously, dipping should continue, but the manner in which it should continue would require further investigation with all participants within the livestock industry.

8. The role of government in the provision of animal health services

The political system of South Africa that changed in 1994 also led to a profound debate on the role that the state should play. In this process of changes in roles and responsibilities, one frequently asked question is which institution will provide or lead to a more optimal delivery of a particular service: public or private? The provision of veterinary services is one area that is included in these debates. In most production systems the veterinary service covers three broad areas: animal health care, production and public health. According to Gros (1994) animal health services consist of preventative (e.g. tick control) and curative care, as well as the delivery of veterinary pharmaceuticals. Production services, on the other hand, are more specifically geared to increase the production of individual animals and herds. They include services such as artificial insemination to achieve genetic improvements.

Animal health care delivery alone involves a wide range of activities, some of which have private good attributes, while others may best be defined as having public good ones. In the veterinary service area some aspects (i.e. curative services and drug sales) exhibit private good characteristics; thus, other things being equal, if delivered by the private sector an economically optimal quantity is likely to be provided. Curative care is, however, only an imperfect private good. It can be argued, especially where contagious diseases are involved, that all producers benefit when one of their neighbours' animals is cured of a disease. If left untreated, such disease could spread to other animals. Hence there is a spill-over effect of externality in such cases. On the other hand, other aspects such as preventative and promotional services are at best a public good which will likely be under-provided unless undertaken by an entity other than the private sector. Thus, the privatisation of public goods, especially those involving preventative and public health services, is likely to result in significant market failures. The only way such goods can be optimally provided is if some kind of public institution assumes responsibility for service delivery. For instance, although it was a traumatic decision, the killing of animals infected with foot-and-mouth disease in KwaZulu Natal offered some positive externalities, particularly to the other provinces in South Africa and elsewhere. In addition, the vaccination of animals was undoubtedly within public interest since it reduced the risk of disease occurrence.

Until 1998/99, in South Africa's developing areas animal health services (e.g. tick control) were provided at highly subsidised rates or ‘free of charge’. For instance, in the former Venda homeland it costs the government R12.00 and communal farmers R1.00 to dip a head of cattle per year (Randela, 2000).

The abolition of dipping subsidies by the state in the late 1990s deprived both the farmers and the state of the services formerly offered by the diptanks. It is hardly possible to perform diagnostic testing for controlled diseases such as FMD and brucellosis in the absence of the communal dipping set-up. In addition, the costs of performing extension as effectively without this opportunity are enormous and are beginning to render extension unaffordable and unachievable. The principle of ‘pay for services’ is a good one, but not necessarily of first priority in the complicated field of dipping, especially in controlled areas. Currently the South African Government seems to be reluctant to fund animal health services. This may be due to the government's failure to realise the value to the country of the veterinary service or failure to foresee the consequences to society if such a service is not provided.

9. Conclusion

The cost–benefit analysis of ticks and tick-borne diseases control programmes was done in two scenarios: ‘with’ and ‘without’ control. Emphasis was placed on comparing the direct costs and benefits of each strategy at the household level, had both strategies been adopted. The income forgone in terms of milk, meat, draught power and low productivity under the ‘without’ scenario is larger than the income forgone under the ‘with’ scenario.

The results of this paper reveal a cost–benefit ratio which is less than one (0.8) using a 4 per cent mortality rate. This ratio can possibly be improved by reducing the current dipping costs through the adoption of a strategic tick control. However, the ratio seems to be sensitive to changes in the mortality rate, yielding a cost–benefit ratio of 1.2 when the mortality rate is changed from 4 per cent to 10 per cent. Although the cost–benefit ratio is less than one, the dipping programme still deserves government support because of the broader socio-economic service it provides. More importantly, in less developed areas dipping forms part of the whole veterinary surveillance programme, which is in essence a purely public good. A fundamental requirement for the national and international control of animal disease is the provision of a comprehensive system of disease surveillance and disease reporting. The information from the surveillance programme benefits the whole sector and cannot be appropriated by any livestock farmer. Therefore government intervention is necessary to control disease outbreaks and ensure early warning.

The financial support and comments the author received from former colleagues in the Unit for Development Impact Analysis of the Agricultural Research Council are acknowledged with thanks. In addition, he thanks Mr A Spickett for his valuable inputs and comments, particularly in the initial stages of the survey.