Vertical and Spatial Integration of Live animal and Meat Markets in EthiopiaPresentation Transcript
Vertical and SpatialIntegration of Live animal and Meat Markets in Ethiopia By Tadesse Kuma (EDRI) & Seneshaw Beyene (ESSP-II) Paper presented on the ESSP-II/EDRI Workshop on : Taking Stock of the Economy of the Livestock Sector in Ethiopia Jupiter International Hotel, Addis Ababa November 4, 2011
Outline1. Background2. Domestic meat demand3. Brief Description of the Model4. Results and observations5. Summary and Conclusion
I. BackgroundWhy market integrations? Integration of agricultural commodity and rural and urban food markets is a precondition for effective reform ; without integration price signals will not be transmitted; Market based policies for poverty alleviation and food security could be more effective if markets are integrated; Knowledge of market integration also allows monitoring of price movements, forecast prices;
Background (2) Current policy measures and infrastructure development by the government envisaged to bring clear improvement on market integration; Benefiting producers by improving transmission of price signals between markets remained concern; Livestock production and trading are the principal economic activities (30% agr. GDP); Towards this end, this study aims to investigate at 3 levels.
Domestic meat demand (1) Despite the largest livestock population in Africa, Ethiopia’s consumption of animal source food in general remained low.Table 1: Annual per capita meat consumption (kg/hd/year), 2003 Botswana Tanzania Ethiopia Sudan Kenya South Africa EgyptTypeBovine 4 6 9 8 9 8 134Mutton and goat meat 1 1 2 4 7 1 3Poultry & Pork - 1 1 3 0 9 25Total meat (kg/capita) 5 8 12 15 16 18 42Estimated GDP/capita (USD) 346 500 800 7000 1500 2000 5700Cattle equivalent /capita 0.7 0.5 0.4 1.4 1.5 0.1 0.4 Source: FAOStat; World Bank; IMF; Negassa and Jabbar (2008), adopted from GebreMariam et al. (2010).
Domestic meat demand (2) Pronounced differences have been identified between rural and urban patterns of meat consumption (beef: 1.7 kg vs. 7.0 kg/PC/Yr); Annual beef population per head of cattle (i.e., 8kg Eth, 11 in Sudan, 14 in Kenya, 51 in Australia, and 79 in the USA); The current average meat production per cattle (carcass weight: World 212 kg, , Africa156kg, and East Africa stands 143kg/head, Ethiopia 110kg)
Meat export performanceSource: FEWS NET, 2011
The TAR Model – Market Integration (1) Research Objectives: Study there exists vertical integration between producer and retail bull prices, and between retail price of bull and meat price within in selected zones. Check the level and speed of integration, if they are integrated. Models of Market Integration (1) Correlation analysis, Cointegration analysis, PBM( Baulch,1997), Threshold Autoregressive Models
The TAR Model – Market Integration (2) Consider Symmetric assumption (instant adjustment) Ignores presence of transaction costs Doesn’t capture non-linearities due to reversals in direction of trade flow
The TAR Model – Market Integration (3) But it has been noted that: dt-1 > Ө ij, dt-1 <- Өij, and |dt-1|< Өijthe Band [-Ө ij , Ө ij] Where: the price differences should exceed TC (Ө) before initiating an arbitrage for rational traders to engage in arbitrage
The TAR Model – Market Integration (4) out dt 1 t : dt 1 dt in dt 1 t : dt 1 d out t 1 t : d t 1 Ө estimated by a grid search Assumes constant threshold/transaction cost (θ ) TCs (Ө ) could vary depending on season, direction of trade, quality of roads and fuel prices
The TAR Model – Market Integ (5) Following Van Campenhout (2007) The thresholds and adjustment parameters to vary with time, t. out dt 1 .t.dt 1 t : dt 1 t out dt t : t dt 1 t d .t.d out t 1 out t 1 t : dt 1 t Half-life: is the solution for T in dt+T=dt/2 and it is calculated as T= ln (0.5)/ln(1+ ρ )
Results of TAR Analysis 50 zones from 8 regions were originally considered Following requirement by the model for the prices to be non- stationary in level terms, 8 Stationary price (in level terms) series are dropped for VPT analysis: 42 Producer-Retail pairs of zonal markets considered; 36 Retail-Meat pairs zonal price levels were chosen for SPT analysis: 37 Retail – Retail pairs of spatially separated zonal markets,
Non Integrated zonesSummary of zones with no vertical/spatial integration(1). Producer-Retail (2). Retail-Meat (3). Retail – Retail (vertical) (vertical) (Spatial)Afar Zone 03 Afar Zone 03Afat Zone 01 Afar Zone 01North Wollo North Wollo North Shewa East ShewaEast Hararghe East HarargheJimma Jimma Shaka-KafaIllubabor Borena Borena-BaleGamo Gofa Gedio Gedio-SidamaWest Wollega Arsi-Bale Harari South Tigray East Gojam Sourth Gondar
Results of analysis: synthesis Strong spatial integration of AA and Harari with other markets; Weak spatial integration in Somali and Amhara Zones in every region have shown strong vertical integration of producer-Retail markets of ox except those in Amhara Transaction costs seem to decline (in nominal terms) between Sept 1998 and Feb. 2011 in Tigray, Somali, and Benishangul Very week vertical integration between oxen and beef Speed of adjustment is very slow-the fastest being in Harari with 3 months.
Results of analysis: synthesisA. Spatial Integration of oxen markets (Retail – Retail) Thresholds With out time trend With time trend (transaction costs) Regions Coefficient P Hlaf life Coefficient P Initial EndTigray -0.609 0.00 0.763 0.002 0.36 0.23 0.17Afar -0.682 0.00 0.604 0.001 0.50 0.16 0.41Amhara -0.449 0.04 1.404 0.000 0.54 0.21 0.23Oromiya -0.672 0.01 0.728 0.002 0.24 0.17 0.26Somali -0.378 0.07 1.459 0.001 0.71 0.61 0.64Benishangul -0.523 0.01 0.938 0.001 0.59 0.13 0.23SNNPR -0.737 0.00 0.546 0.003 0.31 0.32 0.28AA -0.915 0.00 0.282 0.004 0.03 0.40 0.46Harari -1.040 0.00 0.112 0.005 0.03 0.12 0.21 AA and Harari seem to be well integrated with other markets Weak integration in Somali and Amhara
Results of analysis: synthesisB. Vertical Integration (Producer-Retail) Thresholds With out time trend With time trend (transaction costs) Regions Coefficient P Half- life Coefficient P Initial EndTigray -0.788 0.01 0.609 0.003 0.25 0.19 0.17Amhara -0.585 0.01 0.964 0.001 0.43 0.19 0.22Oromiya -0.827 0.01 0.489 0.003 0.25 0.15 0.19Somali -0.775 0.03 0.498 0.001 0.36 0.18 0.13Benishangul -0.835 0.00 0.371 0.001 0.37 0.15 0.10SNNPR -0.866 0.01 0.440 0.004 0.25 0.19 0.35 Strong vertical integration between Producer and Retail prices except markets in Amhara; Lesser transaction costs (in nominal terms) in February 2011 as compared to Sept.1998 in Markets in Tigray, Somali, and Benishangul
Results of analysis: synthesisC. Vertical Integration (Retail-Meat) With out time trend With time trend Regions Coef. P Half life Coef. PTigray -0.043 0.01 16.528 0.000 0.34Amhara -0.030 0.04 25.073 0.000 0.39Oromiya -0.051 0.04 19.807 0.000 0.40Somali -0.053 0.05 12.726 0.000 0.51Benishangul -0.025 0.03 27.317 0.000 0.62SNNPR -0.084 0.01 10.528 0.001 0.18Harari -0.200 0.00 3.100 0.001 0.00 Very week integration between beef and ox Speed of adjustment is very slow for all markets. The fastest being in Harari with 3 months; No integration in Afar
Summary and conclusion In general, live animal and meat market development is at its infant stage with several weaknesses; Virtually no vertical integration was found for oxen retail and meat markets at regional as well as at zonal level. Speed of adjustment is observed abnormally high; Markets in Afar and Somali regions in most cases are found to be weakly vertically integrated compared to other regions; In most cases transaction cost has declined at the end of the period (March 2011) compared to the initial period (January 1998).
Implications Strengthening market institution, infrastructure and information systems; Strengthening producer and consumers organization; Reducing transaction costs ( mainly transportation)The way forwardSurvey based and detailed field studies on the live animal and meat value chain fromproducer to final consumer covering all market actors (producer, middlemen,retailer, abattoirs, butchers, feed lots, meat processors, exporters, etc.) are requiredto obtain concrete evidence on the performance of the sector.