Introduction
Prices are the most readily available and reliable information that guide farmers’ planting decisions in Nigeria. A farmer’s planting decisions depend on anticipated profits which in fact depends on anticipated prices of planted crops. This has made prices an important tool in the economic analysis of markets.
Market integration refers to the comovement of prices and more generally to the smooth transmission of price signals and information across spatially separated markets. In a developing economy like Nigeria, the dynamics of the exchange of information and its effects on the pricing processes are not well understood. This has made prices the must reliable information source in Nigeria’s agricultural marketing systems.
Most studies of agricultural product prices in Nigeria focused on vertical dynamic analysis (Olayemi, 1977; Oludimu, 1982; Mobolaji, 1977; Adekanye, 1988; Afolami, 1988; Adeyokunnu, 1973; Okunmadewa, 1990; Ladipo and Fabiyi, 1982). Of recent, a growing number of studies on agricultural market integration in Nigeria have focused on different aspects of agricultural marketing (Dittoh, 1994; Mafimisebi, 2000; Oladapo, 2004). This paper will extend the study of market integration in two aspects. Previous studies on market integration focused on single product. The study compares the market integration of different product markets and measures the degree of market integration by using the Index of Market Concentration (IMC). Studies of market integration can help policy makers design appropriate agricultural product supply across the state so as to avoid too much instability in the rural economy.
Study Area
The study area is Oyo State which is one of the thirty six states in Nigeria.
The State is located within latitude 3° and 5° North of the equator
and between longitude 7° and 9° 3^{1} East of the Greenwich
Meridian.
Oyo State covers 35,743 square kilometers and has an estimated population 11.5 million as at 1996. The state consists of thirty three local government area (LGAs).
The State is characterized by equatorial climatic conditions. There are two distinct seasons namely wet and dry seasons. The wet season covers between November to March and is characterized by hot weather. The topography is about 05 m above sea level and the mean annual rainfall is within the range of 10001400 mm (MANR, 1999).
The soil resources of Oyo State may be divided into two broad groups. These are the poorly drained soils and the welldrained soils. The alluvial or poorly drained soils are found on the flood plains of Ofiki, Opeki, Ona, Oba and Sasa rivers. These soils are widely cultivated to grow yam, maize and sugar cane. They are most useful during the dry season when farming activities have stopped on higher grounds. They are three main types of welldrained soils. These soils support most of the farming activities in Oyo State. The crops grown include both annual and perennial crops such as yams, cassava, oranges, cocoa, tobacco, cashew, etc.
The vegetation consists of the savanna woodland and the forestlands. Savanna woodlands covers over half of the land area of Oyo State. It is often divided into the derived savanna and the gallery forests. The only forests found in the state are along river valleys and the called gallery forests.
Data Collection
The data for the study were sourced from the Oyo State Agricultural Development
Programme (OYSADEP). It is one of the MSADP I projects approved by the World
Bank for assistance in Nigeria. Average commodity market price data was collected
through the Planning Monitoring and Evaluation (PME) unit of OYSADEP. For the
price survey, four zonal extension offices with 20 block extension offices served
as contact points for the collection of rural and urban market prices. Average
monthly prices in Naira per kilogram (N/kg) for cassava, tuber, yam, yellow
and white maize were collected for 8 years; 1994 to 2001 for both urban and
rural markets. The urban markets in Oyo State include Bodija, Oje, Gambari,
Ilora, Owode markerts, while the rural markets include Akanran, Towobowo, Anko,
Irepodun, Oje, Obada, Ipapo and Igbeti.
The integration of different product markets and measure the degree of integration by using the Index of Market Connection (IMC). Studies of market integration can help policy makers design appropriate agricultural product supply across the state so as to avoid too much instability in the rural economy.
Model Specification
The Ravallion Model seeks to determine whether a change in the price of
a product in a local market is influenced by the change in the central market.
Ravallion’s approach was used to develop a structural model of prices (Ravallion,
1986). Formation in N local markets by assuming that local prices (P_{1}……,
P_{N}) are dominated by one central or reference market price (R). The
static form of the model can be represented as follows:
Where X_{i} = vector of seasonal or other exogenous variables which might influence price formation in market I and central market.
The dynamic structure of Eq. (1) and (2)
if expressed in a linear form are:
Where:
R_{t} = αR12_{t1} + β_{10}P_{it} + B_{20}P_{2t} +…β_{NO}P_{NE} + β_{11}P_{it1} + β_{N1} P_{Nt1} + CX_{t} + ∈_{t}
(∈_{it} and ∈_{t} are suitable error processes)
Note the followings about the equations
• 
Only one lag of each endogenous variable has been included,
but a general model with ‘n’ lags of local prices and ‘m’
lags of the central price is possible. 
• 
Because of the nature of transport costs, the model was estimated
with actual prices rather than their lags. 
If Eq. (3) is re written in the form of error correction
mechanism, that is using ‘Δ’ for the difference operator.
Δp_{t} = P_{t}P_{t1}. Thus
Δp_{it} = (a_{r}1)(P_{it}
– 1R_{t1}) + b_{i0}ΔR_{t} + (a_{i}
+ b_{i0} + b_{i}11)R_{t1} + C_{i}X_{it}
+ ∈_{it} 
(5) 
Since there is likely to be less collinearity in Eq. (5)
than the equivalent Eq. (3) this error correction form was
estimated. Tests for market segmentation is given by
On the other hand tests for long run integration is indicated by
b_{io} =1, b_{it} = a_{i}
= 0 
Index of Market Connection Analysis
Index of Market Connection (IMC) is used to measure price relationship between
integrated markets and following formula was used to calculate IMC:
P_{t} = β_{0} β_{1}P_{t1}
+ β_{2} (R_{t} – R_{t1}) + β_{3}R_{t1}
+ ∈_{t} 
Where:
R_{t} 
= 
Urban or reference price 
P_{t} 
= 
Rural price 
R_{t1} 
= 
Lagged price for urban markets 
R_{t}R_{t1} 
= 
Difference between urban price and its lag 
∈_{t} 
= 
Error term or unexplained term 
β_{0} 
= 
Constant term 
β_{1} 
= 
Coefficient of rural lagged price 
β_{2} 
= 
Coefficient of RtRt1 
β_{3} 
= 
Coefficient of urban lagged price 
IMC 
= 

According to the model, IMC equals to the coefficient of lagged price in local
markets divided by the coefficient of lagged in reference market interpretation
of the IMC.
IMC 
<1 
implies high shortrun market integration 
IMC 
>1 
implies low shortrun market integration 
IMC 
= ∞ 
implies no market integration 
IMC 
= 1 
high or low short run integration (theoretically) 
The closer the value of the IMC to zero, the higher the degree of market integration and by extension the higher the marketing efficiency. In order to capture the IMC values better, the values were approximated to two decimal places.
Results and Discussion
Table 1 shows the regression results for the market pairs for four crops, cassava, yam, white maize and yellow maize.
The coefficient of multiple determination (R^{2}) shows the percentage of the rural price (P_{t}) that is explained by the lagged rural price (P_{t1}), difference between urban price (R_{t}) and its lag (R_{t1}). The regression equation explained 88.3, 86, 96.6 and 92.5% of all the variabilities in the rural prices of cassava, yam, white maize and yellow maize, respectively. The Durbin Watson test was conducted on the data to detect the existence of serial correlation. The result in Table 1 indicate the nonexistence of serial correlation since the Durbin Waston values we re either approximately equal to two for the crops covered by the study. The Ftest indicates that the regression equation is significant at 10% .
The Indices of Market Connection (IMC)
The indices of market connection (IMC) is used to measure price relationship
between integrated markets. For the cassava, yam, white maize and yellow maize
market pairs the IMCs were 0.3074, 0.0814, 0.02712 and 0.1648, respectively
(Table 2). The IMCs for these market pairs are all less than
unity and very close to zero thus indicating high degree of short run market
integration.
Table 1: 
Regression analysis results for Oyo state 

SourceData Analysis 2004, Figures in parentheses are tvalue
calculated,***  Coefficients significant at the rate of 1% ∝, ** 
Coefficients significant at the rate of 5% ∝, *  Coefficients significant
at the rate of 10% ∝, S/RShort run 
Table 2: 
IMC and classification of markets in Oyo state 

Source: Field data analysis, 2004 
The IMC for white maize indicate a higher degree of market integration than
yellow maize. This may be explained by the high demand for white maize for the
preparation some local foods in the area.
These results confirm that price changes in the urban markets (Bodija and Ilora) immediately cause a price change in the rural markets (Akanran, Towobowo, Anko, Irepodun, Oje, Obada, Ipapo). The high degree of integration in these markets is explained by the short distances between the rural and urban markets and the channel of distribution of these staples. Bodija and Ilora serve as terminal markets for the nearby rural markets covered by the study.
These food crops reach the market from the farm in four principal ways; by means of direct sales to rural and urban consumers; direct sales to rural assemblers, direct sales to retailers and direct sales to terminal markets. Farmers transport these food crops to the terminal markets using pickup trucks over relatively short distances up to 20 km and they sell directly to wholesalers. Farmers also sell small quantities to rural assemblers. These assemblers finally sell to urban based wholesalers who move from one village market to the other; to assemble these products.
Conclusions
Through this analysis of cassava, yam, yellow and white maize market integration, it is concluded that the grains and the tuber markets in Oyo State are highly integrated. Thus, price signals are transmitted from food deficit urban markets to food surplus rural areas. The study did not indicate a fully integrated market (B_{1}≠0) and complete market segmentation did not exist (B_{3}≠0). Using the Indices of Market Connection as a proxy for marketing efficiency, we infer that the grain and tuber markets in Oyo State are highly efficient in the short run, thus the market pairs are not characterized by much market imperfections. This is due to the short distances between the reference and the rural markets as well as the direct interaction of urban wholesalers with the farmers facilitated information flow between reference and rural markets.
Despite the progress in market performance, some inefficiency remain. However these inefficiencies and absence of the necessary storage infrastructures, storage, market information, standardized weights and measure and other market support services still impair further free flow of goods and services.