home.cd3wd.ar.cn.de.en.es.fr.id.it.ph.po.ru.sw Misd010.htm
CATTLE MARKET PRICE AVERAGING AND FORECASTING - TANZANIA 1994
MARKETING INFORMATION SYSTEM (MIS)
INTRODUCTION 1999
This report with graphs was done by me 1994 and 1995.
I am reissuing it because:
The whole job was in several parts:
So I enclose here the Data Collection and Analysis methods and the analysis of historical data; I exclude the computer system at this point in time. Anyone who wants dbase tables of the historical data or an Access database (only 882k when winzipped (TM), 7 megabytes when unzipped) of the historical data just email me on alexweir@usa.net and I can email the stuff to you. Also the Harvard Graphics .cht data files can be supplied for anyone who wants them. I am also always interested in rewriting the data input, processing and reporting system in some modern programming language like VB.
I think this report and the related data and program as above should or may be relevant to:
- Universities and agricultural colleges
The complete system can be run from your hard disc using any browser, and it is also intranet and internet-ready - just drop all the .htm , .gif and .jpg files into one directory and double click on INDEX.HTM to start up the system.
The report comprises 160 graphs (100 with explanation and 60 largely self-explanatory), about 10 photos, and approximately 56 pages of text. Total 231 .htm files @ 2.4k bytes average, plus 160 .gif files for the Harvard Graphics graphs @ 14k average, plus 14 .jpg files for the photos @ 100k average size. Total size is 4.1 megabytes (approx 3 floppy discs) or 3.6 megabytes when zipped using WinZip (TM). Takes about 15 minutes to upload or download using dialup email at 33k baud.
Since my original report in 1994/95 had no Acknowledgements let me do a quick one here:
Dominic Massawe , Sophia, Kadigi and the others at TLMP for their help and cooperation during and after my work there September 1994. Percy Robb, Frank Skov, Verner Jensen, Jan Stryffert, Ole Jepsen and others at Danagro in Copenhagen. All the people we interviewed and who helped otherwise during the Project. Percy Robb I am particularly grateful to towards the end of my initial 7 weeks onsite work by suggesting that the phenomenon we were seeing were more related to cattle liveweight than to cattle grade as such - that led to the log-log relationship between price and edible weight which forms a large (but not complete) part of this report. Jim Airey and Martin Doran without whose initial work and without whose massive data collection effort this report could not have existed. UNFAO and UNDP Dar es Salaam for employing Jim and Martin.
Speaking of which, one does not have to accept that log log methodology in order to gain something or even a lot from this report and from the data contained. That 5 years data itself is worth a goldmine even without my interpretation - it is open to re-interpretation by anyone and everyone.
And by the way we never ever compared the weights data with the rainfalls for those years - that could even produce results which will enable forecasts to be made based on recent rainfall figures.
Weighing was regarded (correctly I think) in this report as a onerous, labour-intensive and problematic activity - now that digital cameras are commonplace and almost affordable, I am sure that they could be used along with PC's for weighing - one shot sideview with 2 metre sticks fastened at 90 degrees held just in front of the animal like a cross; then one shot from the rear with a single metre stick held horizontally again just in front of the animal. Bring up on the computer screen in a suitable software package enables precise measurement of body length, body height and body maximum width. Experimentally-derived formulae enable immediate calculation of liveweight, carcase weight and edible weight to within a few kilogrammes plus or minus - i.e. the system not only calculates liveweight but also automatically grades the animal (based on the body maximum width).
Apologies for not checking through the report and rewriting parts or the whole - there just isnt enough interest guaranteeable in this kind of stuff to justify that. I did rework the colours and backgrounds to the graphs to make them nicer to look at, and I did reinclude some graphs (the 60 without comment) from earlier drafts which seem very relevant. Also the photos of grades and grading and some photos from Moshi cattle market are included.
Anyone who wants more info, data or just to discuss anything related to the Project, please email me on alexweir@usa.net If you are viewing on internet or an intranet, and you want the report for your hard disc, then email me and I can email you the zip file at 3.6 megabytes.
EXECUTIVE SUMMARY 1999 - I dont think I summarised too well in 1994 what this report is about, so I do it again here - In Tanzania there is very little or no auctioning of cattle - all or most sales are done through private deals. Because of that someone (probably correctly) thought that there would be scope for an information system which gave sellers and buyers and idea of what prices would be like at the various markets for the various weights and grades of cattle. Up till 1994 this market weight, grade, sex and price sampling would typically sample 35 cattle per market day. This report works out a way to reduce that to about 6 - 8 cattle, and with increased accuracy. Since the sampling 1989 thru 1994 was REPRESENTATIVE, then this report also collates and graphs lots of interesting and useful data regarding patterns and trends for the 3 major markets - Dar es Salaam, Arusha and Moshi. One of the really interesting things we found was that Arusha awards financial incentive to better grade heavier cattle, but Dar treats all cattle alike in terms of US$/edible kilogram.
Alex Weir, December 1999