As outlined above, there is now an established body of knowledge about the economics of the tsetse and trypanosomiasis problem based on careful field work spanning the continent. Most of this has been very ably summarized in Brent Swallows position paper for PAAT (PAAT, 2000), which concentrates on attempts to quantify the impact of the disease in economic terms. In particular, this paper provides a summary of the data that has been collected in the course of studies over the past two decades about the effects of trypanosomiasis on cattle production parameters (mortality, fertility, milk yield and draught performance) and discusses the methods used to collect this data. Two earlier reviews (Itty, 1991; Tacher et al., 1998) included more of the earlier studies and also covered benefit - cost studies. The most significant study to appear since Swallows review is the socio-economic volume of the three-volume study produced in 2000 by the Regional Tsetse and Trypanosomiasis Control Project (RTTCP). This volume (Doran, 2000) is based on socioeconomic surveys and studies done in all four RTTCP countries, and concludes: The evidence available from these studies suggests that the continued use of trypanocides to control the disease is the most sustainable and transferable option available at present for Malawi, Mozambique and Zambia. In Zimbabwe, the logical strategy is to hold the target barrier which prevents the re-invasion of tsetse flies from neighbouring countries. The socio-economic study is based on a cost-effective and reproducible methodology and the survey questionnaires are included.
To the continent-wide studies reviewed by Swallow (Jahnke et al., 1988; Kristjanson et al., 1999; Gilbert et al., 2001) is added Budds analysis of the potential benefits of tsetse eradication (Budd, 1999). This built on the suggestion made in Swallows paper (PAAT, 2000) of using estimated elasticities which measure the change in total agricultural output occurring in response to a change in livestock numbers. This was used as a basis for capturing the likely indirect benefits to agriculture from the increase in cattle numbers that would be possible in a tsetse-free Africa. The other insight provided by Budd (1999) was an investigation of the economies of scale arising from very large tsetse control schemes due, not just to the usual factors that make for economies of scale, but also to an effect specific to tsetse clearance, that of needing less expenditure per km2 on barriers to protect from re-invasion. Once the benefits were taken into account, the effect of these economies of scale was that the highest benefit - cost ratios were achieved, on the one hand by the very large projects (due to economies of scale), and on the other by the very small projects (which targeted problem areas and had high cattle population densities). The use of GIS systems (PAAT, in prep.), and specifically of PAATIS has made it possible to be more precise about the potential increase in cattle populations which might follow from the removal of tsetse flies (Gilbert et al., 2001).
These continent-wide studies clearly provide support for the many statements made over the years about the diseases impact and its importance in constraining development (e.g. MacLennan, 1980; Jordan, 1986; PAAT, 2000). This view is reinforced by the results of most of the project evaluations undertaken which have included a benefit - cost analysis. A full review of the many benefit - cost analyses undertaken is outside the scope of this paper. To list just a few (Putt et al., 1980; Brandl, 1988a; Putt et al., 1989; Shaw, Zessin and Münstermann, 1994; Blanc, Le Gall and Cuisance, 1995; Woudyalew et al., 1999; RTTCP, 1999), these typically come up with benefit - cost ratios between 2:1 and 5:1, using discount rates of 10 percent or more (see the section Time value of money, page 11). The studies undertaken have included both appraisals or ex ante studies, which look ahead to try and estimate the potential benefits and/or costs of a project and evaluations or ex post studies, which examine the profitability of work that has already been undertaken. The continent-wide studies are, of necessity, all ex ante estimations of what the effects of controlling the disease throughout Africa might be.
The relative cost-effectiveness of the many approaches towards controlling tsetse and trypanosomiasis has also been looked at for many situations, the most comprehensive study to date being Barrett (1997). The costing out of different approaches towards controlling the disease has also been incorporated into most of the benefit - cost studies cited above. Thus, a hierarchy of costs has evolved - how this can be interpreted is discussed in Chapter 4, in the section Cost side considerations, page 27.
We are, thus, in the happy position of trying to control a disease that everyone agrees is sufficiently important for virtually any intervention to be beneficial. However, there are some clouds on this horizon. First, most writers on the subject have identified situations where certain interventions are economically unjustifiable (and often also technically infeasible), for example tsetse control is less profitable where cattle and human populations are very low, since there are few units to build benefits on and keeping such an area free of tsetse is relatively more expensive. Second, the situation is changing over time, with expanding human and animal populations reducing the flys habitat. This means that in some areas the problem is self-limiting, as originally stated by Nash (1948) and discussed in Bourn et al. (2001) and Reid et al. (1999). Third, the farmers themselves are not passive players in this game and they have developed strategies for dealing with the disease; it is only recently that we have achieved a better understanding of what these strategies are (Pokou, Swallow and Kamuanga, 1998; CIRDES, ILRI and ITC, 2000; Doran, 2000; Kamuanga et al., 2001a).
What is needed, then, is a generally accepted and consistently applied methodology for assessing the economics of controlling the tsetse and trypanosomiasis problem, with a view to establishing an economic filter or screening process, which, in an analogous manner to the GIS filters:
makes it possible to reject control options or areas where interventions are not profitable;
identifies situations where interventions could be profitable but depend on a number of uncertain factors; and
flags up those situations where it is clear that a certain control strategy is highly cost-effective and has significant benefits to offer to livestock keepers and farmers - these would be the priority areas.