




































African Journal of Agricultural Marketing Vol. 1 (1), pp. 011-015, November, 2013. Available online at 
www.internationalscholarsjournals.org © International Scholars Journals 

 

 

 

Full Length Research Paper 
 

Farm household level impacts of information 
communication technology (ICT)-based 
agricultural market information in Ghana 

 
Ramatu M. Al-Hassan*, Irene S. Egyir and James Abakah 

 
Department of Agricultural Economics and Agribusiness, University of Ghana, Legon, Ghana. 

 
Accepted 13 March, 2013 

 
This paper assesses the impact of the information communication technology (ICT)-based market 
information service (MIS) on farm households in the eastern corridor of Northern Ghana. Data was 
collected from 346 farm households in the eastern corridor of Northern Ghana where an ICT-based 
market information project was implemented between 2006 and 2009. The sample included 159 
participants and 187 non-participants. The logit model was used to identify factors which influence the 
probability of participation in the ICT market information project, while propensity score matching 
(PSM) was used to assess the project impact. Results indicate that age and value of assets were 
negatively related to the likelihood of participation in ICT-based MIS. A unit increase in age reduced 
likelihood of participation by 0.5%, while a unit increase in value of assets reduces likelihood of 
participation by 7%. Previous participation in a development project increased likelihood of 
participation by more than 60%. Participation in the ICT-based project increased expenditure on 
pesticides, and food security by 11%. More participants (13%) used improved seed than non-
participants. It is recommended that younger persons are made the target/focus group for ICT projects 
and that the private and public sectors collaborate to facilitate wide availability and expanded use of 
ICT MIS. 

 
Key words: Information communication technology (ICT), agricultural market information, smallholders, Ghana. 

 

 
INTRODUCTION 
 
Development of the agricultural sector in many African 
countries hinges on the development of the smallholder 
systems that have sustained African agriculture to date 
but continue to face challenges of low productivity and 
limited access to remunerative markets. Non-competitive 
value chains and limited information about remunerative 
markets, and risk aversion of smallholders limit their 
integration into markets.  

Ghana’s  agricultural  growth  has  been  erratic  and  at 

 
 
 
 
 
5.2%, it has lagged behind Gross Domestic Product 
(GDP) growth of 5.9% [Institute of Statistical Social and 
Economic research (ISSER), 2005 - 2010]. Ghana’s 
development strategies since 2001 have sought to turn 
the agricultural sector around through modernisation and 
commercialisation of smallholder agriculture (Republic of 
Ghana, 2003a, 2006). The Food and Agriculture policy 
has as one of its key objectives, integration of 
smallholders into domestic and international markets 

 
 
*Corresponding author. E-mail: aramatu@ug.edu.gh, ramatu_mahama@yahoo.com. Tel: +233 20 8154570.  



Al-Hassan  et al.               011 
 
 

 
Republic of Ghana, 2007) with improvement in access to 
market information as a key strategy. The information 
communication technology (ICT) for Accelerated 
Development (Republic of Ghana, 2003b) policy also 
aims among others, to facilitate the modernization of the 
agricultural sector through the deployment and 
exploitation of ICTs to improve the sector’s efficiency and 
productivity. Initiatives taken to improve investments in 
ICT deployment include the privatisation of the 
telecommunications industry, establishment of a 
regulatory authority, and investments in infrastructure 
such as an internet backbone, and masts for mobile 
phones. Community Information Centres (CICs) have 
also been provided in several districts to facilitate training 
in ICT on a wide scale.  

The focus on ICT-based methods of information 
provision is driven by the role they can play in 
communicating knowledge and information to rural 
farmers. Access to market information can enhance 
farmers’ access to markets through better negotiation and 
meeting the demands of the market (Barrett, 2008; Moser 
et al., 2005), on condition that constraints to access to 
inputs are addressed. The ICT tools have included 
modern tools such as the internet, mobile telephony, and 
interactive video and CD-ROM programs, as well as 
traditional ICTs of the radio and television (Munyua, 
2007).  

There has been an increase in the use of ICT 
applications in rural Ghana in response to the enhanced 
policy environment, although issues of availability of the 
ICTs, electricity, literacy, telecommunications and content 
are still prevalent (Sampong et al., 2007; Alemna and 
Sam, 2006). An ICT-based agricultural market 
information service (MIS) was introduced by the private 
sector [Busy Lab through its TradeNet (now Esoko) 
platform]. In 2006, the Sustainable Enterprise 
Development Foundation (SEND Foundation), a non-
governmental organisation operating in the north-east of 
Ghana, adopted the TradeNet market information 
platform to access market information for smallholder 
farmers. Specifically, farmers were trained to access and 
provide market information through text message alerts 
using cellular phones. Prior to this intervention, SEND 
Foundation had been promoting agricultural activities and 
cooperative credit unions with the goal of enhancing food 
security in the area. The objective of introducing the ICT-
based MIS was to link the smallholder farmers to markets 
and promote their commercialisation process. The project 
was implemented for 3 years between 2006 and 2009. 
Initiatives such as this are widespread in several African 
countries (for example, Kenya Agricultural Commodity 
Exchange (KACE) and DRUMNET in Kenya; Malawi  
Agriculture Commodity Exchange (MACE) in Malawi, Market 
Information Systems and Traders’ Organisation in West 
Africa (MISTOWA) (Tollens, 2006; Munyua, 2007; Okello 
and Ndiragu, 2009).  

However, information on the impacts of the initiatives 
on smallholders and the markets they operate in is rather 

 
 
 

 
sparse. This paper assesses impact of the ICT-based 
MIS implemented in north-east of Ghana. Specifically, it 
identifies factors which influenced participation in the ICT-
based market information project, and assesses the 
impact of the use of the ICT-based MIS on the level of 
commercialisation and food security, which are the SEND 
project goals. 
 

 
METHODOLOGY 
 
Methods of analysis 
 
The decision to participate is modelled as a random utility function. 
The farmer decides to participate if the utility derived from 

participation exceeds that from not participating. That is, U
a
 > U

b
; 

where a denotes participation and b denotes non-participation. U
a
 

and U
b
 are modelled as: 

 
U

a
 = x’βa + εa  and U

b
 = x’βb + εb 

 
Since U

a
 and U

b
 are latent, it is the probability of the observed 

decision (participate or do not participate) that are modelled: 
Let Y* be the binary response variable, Yi*Є{0, 1}, i = 1,…,n farm 
households and P is the probability that Yi*=1 given X; 
 
For P(Yi*=1| X) = Prob (U

a
 > U

b
) 

= Prob[x’βa + εa - x’βb - εb > 0 | x]  

= Prob[x’(βa - βb) + εa- εb > 0]  
= Prob[x’β + ε > 0]  

 
The above binary choice model is estimated as a logistic model 
specified as follows: 

Log_ 
P 

_ = ∝ + βj Xj + _  
 

1−P (1)  
   

 
Where ____

_
_ is the odds of Y* = 1 given that Xj and βj is a vector of 

regression coefficients. βj for variable, Xj is the log-odds and the 
exponential form of βj is the multiplicative change in the odds for Y* 

= 1 when the j
th

 variable increases by one unit, holding all other 
variables constant (Finger and El Benni, 2011). This model is used 
to explain farmers’ decision to participation in the ICT-based market 
information system in the context of adoption.  

Following the adoption literature, decisions to participate in an 
innovation programme depend on attributes of the innovation, 
channels of communicating the innovation and extent of promotion, 
as well as social systems (Rogers, 2003). In this paper, the 
explanatory factors are hypothesized to be observable personal 
characteristics of the farmer or their household which set the social 
context (for example, education, age and gender), awareness of 
ICT-based market information project (determined by promotion 
and communication), capital endowment, and level of output 
commercialization (indicator of economic empowerment as well as 
need for the innovation). Membership in a producer organization, 
and number of crop enterprises are also indicators of capital 
endowment. Being a member of producer organization is included 
because it creates opportunity for social interaction which the phone 
can facilitate. The radio and TV are substitutes for the mobile phone 
as a means of accessing information but they may also expose 
potential users, through marketing, to the capabilities of the mobile 
phone. Previous participation in an agriculture development project 
is both a measure of social capital and the farmer’s appreciation for 
social interaction. Distance to main market is included to control for 
the trading environment, which influences need for information and 
therefore likelihood of participating in the 

   



012       Afr. J. Agric. Mark. 
 

 
 
Table 1. Explanatory variables in Logit model participation in ICT-based MIS project. 
 

Variable Description Expected sign 
 

of coefficient  

  
  

Age  
Gender  
Education 

 
Participation 

 
Awareness  
Radio  
TV 

 
Radio/TV 

 
Land 

 
Member of FBO 
 

 
Past agricultural program 
 

 
Needed credit 
 
 
Enterprises 
Sold-out  
Distance to local market  
Distance to nearest phone service 

Asset 

 
Age of farmer (number of years)  
Gender dummy (1 if farmer is a male and 0 otherwise)  
Number of years of schooling of farmer 

 
Participation in ECAMIC project (1 if farmer is a participant and 0 
otherwise) 

 
Awareness of the ECAMIC project (1 if farmer is aware and 0 

otherwise) Radio dummy (1 if farmer had a radio and 0 otherwise)  
Television dummy (1 if farmer had a television and 0 otherwise) 

 
Radio and TV dummy (1 if farmer had a radio and/or TV and 0 
otherwise) 
 
Size of land cultivated before ECAMIC project inception (acres) 

 
Dummy for FBO membership (1 if farmer hold any membership a 
farmer-based organization and 0 otherwise) 

 
Ever participated in an agricultural program in the last 5 years (1 if 
farmer has ever participated and 0 otherwise) 

 
Dummy for need for credit (1 if farmer needed credit for an agricultural 
activity and 0 otherwise) 

 
Number of crops produced by the farmer 

(number) Proportion of total output sold  
Distance (in kilometers) to nearest local market (natural logged) 

Distance (in kilometers) to nearest mobile phone service (natural 

logged) Value of farmer’s assets (natural logged) 

 
- 

 
+ 

 
 
 
 
 
 
 
 
 
 
+/- 

 
- 

 
 
 
 
 
+ 

 

 
+ 

 
+ 

 
+ 

 
+ 

 

 
project to access information. Although owning the cell phone is not 
a requirement for participation, those who own mobile phones are 
more likely to participate in the ICT-based market information 
project; the reverse is true because in both situations the mobile 
phone is the tool for accessing the information. The explanatory 
variables for participation in the ICT-based MIS project are 
summarized in Table 1.  

The impact of participation in the ICT-based MIS project is 
assessed via the method of propensity score matching (PSM), 
which is used to correct selection bias in the assessment of impact 
based on a control group of non-participants (Becerril and Abdulai, 
2009; Ali and Abdulai, 2010). The aim of matching is to line-up 
comparison individuals to sufficient observable factors to remove 
systematic differences in the outcome between treated and non-
treated (Blundell and Costa Dias, 2008; Deheja and Wahba, 2002). 
PSM takes two individuals that are exactly similar in all 
characteristics except the treatment (in this case, participation in 
ECAMIC) and computes the difference in the outcome between 
them.  

The outcome variables used in this study are, income from crop 
sales, the extent of commercialisation, and level of food security. 
Household commercialisation is measured by the share of crop 
output sold; this is a partial measure that captures 

 

 
commercialisation on the output side only. Food security is 
measured by the number of months the farmers’ major staple last 
through the year and also a dummy variable with 1 if the household 
calorie intake per person meets the recommended daily allowance.  

In this study, the observable characteristics by which participants 
and non-participants are paired are age, education, value of assets 
and distance to nearest local market and distance to centres with 
electricity. Kernel based matching (KBM) of participants and non-
participants are used. 

 
Data 
 
The data is from a sample survey of 346 households, using a 
structured questionnaire. The study area was stratified into project 
and non-project communities. A total of 159 beneficiary and 112 
non-beneficiary respondents were selected randomly from 
beneficiary communities. The remaining 75 households were 
randomly selected from non-project communities.  

Household level data included household characteristics, 
awareness of existence of ICT-based MIS in the locality, number of 
ICT-based MIS used, other information sources used, participation 
in farmer organizations, access to formal and informal financial 



Al-Hassan  et al.               013 
 
 
 

Table 2. Estimates of logit model explaining participation. 
 

 Variable Coefficient S.E z p-value Marginal effects (dy/dx) 
 Age -0.04 0.02 -2.11 0.035** -0.005 
 Youth 0.51 1.34 0.38 0.704 0.057 
 Education 0.03 0.05 0.52 0.601 0.002 
 Household size 0.13 0.11 1.18 0.239 0.014 
 Radio/TV 0.54 0.54 1.00 0.319 0.065 
 Land 0.10 0.35 0.31 0.755 0.012 
 Needed credit 0.70 0.63 1.11 0.265 0.093 
 Distance to local market -0.64 0.31 -2.11 0.035** -0.072 
 Past agricultural program 4.01 0.57 7.00 0.000*** 0.617 
 Asset -0.73 0.25 -2.94 0.003*** -0.081 
 Constant 3.63 1.67 2.16 0.031  

 
Dependent variable = Participation (dummy = 1 if a participant); Number of observation = 213; Pseudo R

2
 = 0.43; Log likelihood = -68.43; LR 

Chi
2
(9) = 104.32; Prob > Chi

2
 = 0.0000; S.E = Standard error; M.E = Marginal effects. 

 

 
services, asset levels, crop and livestock production, participation in 
input and output markets, the types of markets used, and sources 
of income. 
 

 
RESULTS AND DISCUSSION 
 
Participation in ICT-based MIS project 
 
The analysis of determinants of participation was 
conducted on the 187 households in beneficiary 
communities, who were aware of the project and 
therefore could choose to participate or not. Participants 
who were aware of the project were identified as ex-post 
(after the sampling) through their response to a question 
in the questionnaire, which asked if the respondent was 
aware of the project or not. Table 2 presents results of 
the logit model estimation for the identification of factors 
that influenced participation in the ICT-based market 
information intervention. The likelihood of participation 
decreased by 0.5 and 7%, respectively, for a unit 
increase in age and distance to a local market. A unit 
increase in the value of assets decreased the likelihood 
of participation by 8%. Those who had participated in an 
agricultural project were 61% more likely to participate in 
the MIS project. There is therefore a tendency for 
repeated participation in projects. This suggests a 
multiplicity of projects targeting same beneficiaries, and 
likely creation of dependency of some farmers on 
external interventions.  

Contrary to expectations, farmers who accessed distant 
markets were less likely to participate in the project. 
Ordinarily one expects these farmers to have need for 
information; however the conduct of markets in the study 
area is such that those who sell or buy in distant markets 
usually have established trading partners (commonly 
called ‘customers’) and therefore will have less need for 
an MIS. Participation in such projects also demands time, 
and distant traders who are more likely to be frequently 

 

 
absent from the community are not likely to participate in 
the projects. 
 
 
Impact of participation in ICT-market information 
project on smallholder farmers 
 
The effect of participation in the ICT-MIS project on 
smallholder farmers was estimated with the KBM method 
and the results are presented in Table 3. The clustering 
of participants and non-participants in the propensity 
score range of 0.4 and 0.7 (Appendix 1) shows good 
matching based on the observed characteristics.  

Thirteen percent more of participants than non-
participants used improved seed and this was significant 
at 1%. More participants than non-participants also 
attained higher level of food security indicators of 
Recommended Daily Allowance (RDA) of caloric intake, 
for example, at the higher level of indicator based on 
IFPRI’s RDA. Participants also spent more on pesticides 
but the level of significance was at 10% only. Since the 
use of pesticides is linked to growing of maize, it 
suggests that participants were growing more maize than 
non-participants. 
 
 
Conclusions 
 
This paper assessed the factors influencing the likelihood 
of participation in an ICT-based MIS, and the impact on 
farm households.  

Participants in the ICT-based MIS project tended to be 
younger than non-participants and slightly more literate. 
The current study in Northern Ghana shows that 
participants of the ECAMIC project used more improved 
seed, spent more on pesticides and attained higher levels 
of food security. The enhanced access to market 
information has increased their orientation to produce for 
the market. Since the analysis explicitly considered the 

   



014       Afr. J. Agric. Mark. 
 

 
 
Table 3. Average treatment effect on treated: effect of participation. 
 

Outcome variable ATT Critical level of hidden bias (г) Number treated Number control 
Commercialization index 0.03(1.02) 2.90 - 2.95 159 187 
Food secure in maize 0.05(0.09) 3.30 - 3.35 159 187 
Food secure (100% IFPRI RDA) 0.11(2.01)** 5.00 - 5.05 159 187 
Food secure (80% IFPRI RDA) 0.09(1.54) 3.50 - 3.55 159 187 
Food secure (100% WFP/UNHCR RDA) 0.13(2.12)** 2.80 - 2.85 159 187 
Food secure (80% WFP/UNHCR RDA) 0.10(1.58) 2.20 - 2.25 159 187 
Crop income 349.89(0.23) 3.15 - 3.20 159 187 
Value of pesticides 1.38(1.71)* 3.90 - 3.95 159 187 
Value of fertilizer -1.85(-0.18) 2.25 - 2.30 159 187 
Use of improved seed 0.13(2.80)*** 5.65 - 5.70 159 187 
Store for future sale 0.01(0.30) 1.00 - 1.05 159 187 

 
Values in parentheses are t-values. ATT is the average treatment effect for the treated. *, ** and *** denote significance of t-statistics of mean 
difference at 10, 5 and 1%, respectively. 

 

 
causal relationship between participation in ICT-MIS and 
household adoption of improved farm technology (seed, 
fertilizer and pesticides), it addressed the counterfactual 
questions that may be significant in predicting the impacts 
of policy change. 
 
 
Policy implications 

 
International Development Research Centre (IDRC) to 
conduct the research project on the Role of ICTs in 
Linking Farmers to Markets in Africa. We also extend 
appreciation to all team members of the eARN project 
especially Dr Julius Okello and members of the advisory 
council for their counsel and support in various ways. 
 

 
REFERENCES  

The policy implication of the findings is that promotion of 
high technology such as ICTs for communication with 
farmers should be targeted to youth in agriculture 
because they are already familiar with the communication 
tools and require little additional training for their use. 
Such programmes or interventions should be supported 
with complementary services such as financial services, 
because project participants tend to have multiple needs. 
In addition, access to production inputs should be 
improved because the use of such inputs is the route to 
commercialization and the ultimate realization of the 
benefits of enhanced market information.  

There is a latent potential for ICTs, especially the 
mobile phone to facilitate transactions of rural farm 
households, which can be realized through positive 
partnership between the private and the public sectors; 
the latter has the responsibility to identify the needs of 
rural farm households that can be addressed most 
effectively by modern ICTs, while the former’s role is to 
innovate to deliver the services cost effectively. Policies 
that improve farm level adoption of improved technology 
and facilities, and road surface conditions and networks 
will be needed to ensure that the farmers can produce 
more (through higher use of inputs and yields) and 
increase sales in more lucrative markets. 
 
 
ACKNOWLEDGEMENT 
 
Authors acknowledge the funding support provided by the 

 
Alemna AA, Sam J (2006). Critical issues in information and 

communication technologies for rural development in Ghana. Inf. 
Dev. 22:236-260.  

Ali A, Abdulai A (2010). The adoption of genetically modified cotton and 
poverty reduction in Pakistan. J. Agric. Econ. 61:175-192.  

Barrett  C  (2008).  Smallholder  market  participation:  Concepts  and 
evidence from eastern and southern Africa. Food Policy 34:299-317 

Becerril J, Abdulai A (2009). The Impact of Improved Maize varieties on  
poverty in Mexico: A propensity score matching approach. World 
Dev. 38(7):1024-1035.  

Blundell R, Costa Dias M (2008). Alternative approaches to evaluation 
in empirical Microeconomics. J. Hum. Resour. 44(3):565-640.  

Deheja R, Wahba S (2002). Propensity score matching methods for 
non-experimental causal studies. Rev. Econs. Stat. 84:151-161.  

Finger R, El Benni N (2011). Farmers’ Adoption of Extensive Wheat 
Production - Determinants and Implications. Paper prepared for the 
122nd EAAE Seminar, "Evidence-Based Agricultural and Rural 
Policy Making: Methodological and Empirical Challenges of Policy 
Evaluation". Ancona, February 17-18, 2011.  

Moser C, Barrett C, Minten B (2005). Missed opportunities and missing 
markets: Spatial-temporal arbitrage of rice in Madagascar. Selected 
paper presented at the American Agricultural Economists Association 
Annual meeting, Providence, Rhode Island, July 24-27, 2005.  

Munyua H (2007). ICTs and small scale agriculture in Africa: a scoping 
study. Draft Report 1 Submitted to International Development 
Research Center.  

Okello JJ, Ndirangu LK (2009). Background study of the status of ICT 
use in Kenyan agriculture. A paper presented at the proposal 
development workshop on “The role of ICT-based interventions in 
linking farmers to markets”, Antananarivo, July 2008.  

Republic of Ghana (2003a). The Ghana ICT for Accelerated 
Development (ICT4AD) Policy. Ministry of Communication, Accra.  

Republic of Ghana (2003b). Ghana Poverty Reduction Strategy. 
National Development Planning Commission, Accra.  

Republic of Ghana (2006). Growth Poverty Reduction Strategy. 
National Development Planning Commission, Accra. 



Al-Hassan  et al.               015 
 
 
 
 
Rogers EM (2003). Diffusion of Innovations, 5

th
 edition. Free Press, 

New York, U.S.A.  
Sampong DD, Irene Egyir S, Osei-Asare Y (2007). Assessing rural 

women food producers capacity to adopt modern ICTS- A case study 

of the Mfantsiman District of Ghana. 2
nd

 Biennial Conference of 
African Association of Agricultural Economists (AAAE), La Palm 
Beach Hotel, Accra August 20-23, 2007. 

Tollens EF (2006). Market information systems in sub-Sahara Africa 
challenges and opportunities. Poster paper prepared for presentation at 
the International Association of Agricultural Economists Conference, 
Gold Coast, Australia August12-18, 2006. 

 
 

 
 
 
 
 
 
 
 
 
 
 
APPENDIX 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 
 
 
 
 
 
 
 



  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
0 .2 .4 .6 .8 1 

  Propensity Score   

  Untreated  Treated  
 

Appendix 1. Distribution of propensity scores. 
 


