23. In Zimbabwe, the cost of production approach is used to set the absolute price levels for beef. However, this approach cannot capture the various dynamic adjustments in input levels, number of cattle and technological variables resulting from a given change in the price of beef. As a result of such deficiency, we examine the feasibility of using price elasticity parameters as an alternative quantitative basis for estimating the appropriate producer price (appropriate in terms of government policy objectives).
24. As mentioned earlier beef pricing policy objectives in Zimbabwe are aimed at attaining self-sufficiency in beef products and at generating a stable flow of foreign exchange earnings from beef exports. When the price of beef rises in the world market, the government would ideally want to see domestic producers exporting the corresponding amounts of beef. This could even be at the expense of domestic consumers, if government attaches a higher premium to the foreign exchange earning objective. In the light of the earlier discussion on the short-run responses of commercial producers to increased beef prices, the quantity of exports is unlikely to increase and may even decrease. Furthermore, since in the event domestic producers will decide to increase their herds in anticipation of further price increases, the domestic supply of beef to consumers can decrease abnormally. As a consequence retail prices will be pushed up.
25. In the case of Zimbabwe consumers, it has been estimated that a drop in the availability of beef per capita of say 5% can result in an upward pressure on retail prices of as much as 10% (Rodriguez, 1985). Adjustments in the CSC prices take a longer process than those in the free market because of the administrative steps involved (including the bargaining process among those with interest in the sector - e.g. farmers, butchers) in determining the new price levels by government. There is thus a possibility of the development of a black market for beef as a result of the negative supply response by producers in the short-run.
26. Comparing the short-run response of Zimbabwe's commercial cattle sector to beef price changes with other countries indicates a wide diversity. For a 10% rise in beef prices, producers in Brazil will cut slaughter levels by 1.1 to 5.6%; Argentina by 6.7 to 9.6%; and Colombia by .58 to 12%. Even providing allowances for technological differences among countries, the dispersion of the absolute elasticity estimates is wide.
27. Beef supply responses to price increases in Zimbabwe have been estimated using different models and independent variables. The first of these, with undeflated beef prices and time trend as independent variables, estimates the beef price expectation behaviour of commercial beef producers through a polynomial weighting function of price (almon model). The result shows that the short-run price elasticity of beef ranged from -0.49 to-0.61 which in other words means that if CSC increases the producer price by 10%, the number of cattle supplied for slaughter will decrease by 5 to 6%.
28. A second (almon) model used average CSC cattle prices deflated by a cost of living index to discount the producer price for the rate of inflation. These prices were further lagged by three to four periods to represent the set of price information known to the producer when he makes beef production decisions. Simply illustrated, this means that a commercial producer will take into consideration the 1986, 1985, 1984 and 1983 cattle prices in deciding on the number of animals he will deliver to CSC for slaughter in 1986. The estimated elasticities and the resulting output adjustments are shown under the first two main column headings of Table 1.
29. The third model is based on the assumption that beef cattle producers attach geometrically declining weights to beef prices of differing "age" - i.e. higher values to newer sets of price information (e.g 1986 prices) versus older sets.
30. On this basis, beef price elasticites were estimated using Zimbabwe commercial farm slaughter functions. Table 1 below presents these estimates per time period and the corresponding percentage adjustments in the number of cattle slaughtered for a 10% rise in the (deflated) price of beef.
Table 1. Estimated beef price elasticities classified by model structure
|
Period
|
Almon Model (Three-period price lags) |
Almon Model (Four-period price lags) |
Geometric Model |
|||
|
Elasticities |
Output Adjustment % |
Elasticities |
Output Adjustment % |
Elasticities |
Output Adjustment % |
|
|
Present |
1.21 |
12.1 |
.15 |
1.5 |
-.44 |
-4.4 |
|
Previous period |
-.95 |
-9.5 |
.24 |
2.4 |
-.28 |
-2.8 |
|
Previous 2 period |
-.48 |
-4.8 |
.42 |
4.2 |
-.17 |
-1.7 |
|
Previous 3 period |
2.60 |
26.0 |
.70 |
7.0 |
-.11 |
-1.1 |
|
Previous 4 period |
N.A. |
N.A. |
1.08 |
10.8 |
-.07 |
-.7 |
|
"Long-run" |
2.38 |
23.8 |
2.59 |
25.9 |
-1.19 |
-11.9 |
Notes:(1) N.A. = Not applicable.
(2) All elasticities estimated at the means for the almon models.
31. The elasticity estimates in Table I represent the potential diversity in the nature of economic reports reaching the key policy-makers resulting from the use of different estimation models. For example, if the cabinet decided to raise real beef producer prices by 10%, an economist using a three-period price lag almon model will forecast an incremental increase in slaughter levels in the current period 8 times higher than the one using a four-period lagged price. Use of the geometric model will result in a prediction of a current 4.4% incremental decline in the number of cattle slaughtered. The four-period price lag model also indicates a continuous increase in slaughter levels for a given beef price increase in al I periods while the geometric model shows the reverse pattern. The almon estimate indicates a 10.8% increase in the number of animals slaughtered in response to the fourth period lagged price while the geometric model forecasts a 0.7% decline.
32. Long-run supply response estimates given in the last row of Table 1 indicate marginal increases in slaughter levels of about 24 - 26% under the almon models, but marginal decreases of about 12% in the geometric model. Conversely, these imply a decrease or an increase of the same magnitude in the number of cattle retained under the almon and geometric models respectively. Hence, if degradation pressures arising from increasing animal populations in relation to beef pricing policies are being scrutinized, then the estimates under the geometric model can lead to economic measures designed to control animal numbers which graze communal areas.
33. The policy signals deriving from these estimates using the different empirical models can be confusing. There are several technical reasons for the unstable signs and absolute magnitudes of the beef price response parameter for the two models. There is no simple way to explain in this paper the technical (statistical) complexities involved. However, much of the problem usually arises from the non-inclusion of some potentially powerful explanatory variables, due to data limitations, in the process of estimating such supply relationships - e.g. livestock research.
34. Such sources of technical problems cannot be totally eliminated from statistical data sets and are particularly acute in African countries. In the light of this, the question becomes: to what extent can policy-makers place their faith in quantitative estimates? Most decision-makers burdened with myriad responsibilities may not even afford the time to judge the strength of the basis of their decision. For the die-hard quantitative supporters (a likely rare specie of policy-makers?) it is quite crucial to validate the models used in computing the critical economic parameters. Validation involves not only subjecting the models to rigorous statistical tests but also requires the subjective evaluation of industry specialists.
