On the road to salvaging livestock market data

Digitizing livestock market data from Kenya’s arid and semi-arid regions

Since independence, Kenya has grappled with collection of livestock market data. The livestock production department was formed in 1987 and since then, the ministries of agriculture and livestock have been split and merged several times, negatively impacting the livestock sub-sector. The creation and dissolution of ministries has come with a change in priorities, which has resulted in glaring gaps in collection of livestock market information. Access to information on livestock prices, volumes traded at markets, breeds, sex and age group of animals, has been hampered. Previous efforts to close these information gaps have not borne fruit due to intermittent funding from the National Treasury. Recurrent droughts in the recent past have compelled government and development partners to embark on efforts to boost resilience of pastoral communities to secure their livelihoods. To achieve this goal, it is important to have information that will help producers make better choices in livestock trade. This need has led the State Department of Livestock, through World Bank’s Regional Pastoral Livelihoods Resilience Project (RPLRP) to revive data collection across livestock markets in pastoral and agro-pastoral counties of Kenya. Despite this effort, there have been challenges in collecting raw data. In recognising this, the State Department of Livestock (SDL) partnered with the international livestock research institute (ILRI), to salvage livestock market data that has been collected over the years but not available in the SDL database.

About 80% of Kenya’s land mass comprises arid and semi-arid lands (ASALs). This environment lends itself to livestock production, necessitating a renewed focus on the livestock sub-sector. By 2008, about 60% of Kenya’s beef production was from the ASALs employing around 90% of the local population1. Data collection efforts tapered off after the end of USAID’s Global Livestock Collaborative Research Support Project (GLRCP) in 2008. The project developed the Livestock Information Network and Knowledge system that was instrumental in collecting livestock market data.

Since 2016, there has been renewed effort to continue with collection of livestock market data through RPLRP.  We now have data monitors who are drawn from county departments of livestock and partner agencies like Kenya livestock marketing council (KLMC) that are engaged in a process to consolidate and digitize livestock data. Data monitors are field personnel from RPLRP and partner agencies tasked with collection of livestock market data in livestock producing counties of ASALs of Kenya.

The National Livestock Information system (NLMIS) comprises a web portal that aggregates livestock market data such as prices, volumes and grades. Data monitors send aggregated data in form of short message service (SMS) codes to the system using data sheets as shown above. The code has market name, name of data monitor, data of data collection, type of animal, breed of animal, sex, category of the animal, aggregate price of the category and volume of a particular animal type. These codes provide a simpler way of relaying livestock market data to the server at minimal cost.

Objective of the data consolidation exercise

The goal of this exercise was to identify which organisations collected livestock market data between 2003 to present (2018), where they collected it and the status of that data. Additionally, we aimed to obtain as much disaggregated data as possible, from agencies collecting data with the purpose of digitising and providing access of the resultant database to NLMIS for public use.

Sample filled NLMIS data sheet

 Currently NLMIS is operational with data monitors sending SMS codes to the server every week. However, running the data system has not been without challenges. The system experiences technical glitches where messages are sent but don’t reach the server. This means that the data sent is lost and cannot be retrieved. The other drawback is distance between local livestock markets and town centres. Mr. Dara, a data monitor, lives in Hola town but must travel to Bangale market (198Km away in Tana River county) to collect data. He is mostly compelled to use matatus (local passenger service vehicles) that are unreliable because not many people travel between the two towns. Sometimes he uses government motorbikes, but even those are not always fuelled due to unsteady flow of funds to the counties. Without proper facilitation it becomes difficult to go to the markets every week to collect livestock data. As such, every missed week implies lost data that could have been instrumental in helping traders and producers in their daily transactions. This situation obtains in all the five counties we visited (Laikipia, Isiolo, Marsabit, Garissa and Tana River).

Accessibility to some of these markets is greatly hindered by the poor state of roads that are sometimes swept away during heavy rains. On the trip to Bura, sections of the road were completely cut off after heavy rains. We had to meander through the bushes, mostly getting lost along the way before getting back on track. Our driver had to muster a few words in Pokomo and ask for directions. Again, this scenario is replicated in most parts of the counties we visited. This is because they are predominantly semi-arid with loose soils. In the event of heavy rains there is usually a lot of surface runoff where top soil is washed away.

Section of Bura-Garissa road swept away by floods

Why is it important to digitize livestock market data?

The process of data digitisation is critical in ensuring Kenya does not lose livestock market information that has been collected over the years. The value of livestock market information is amplified when it is accessible from a central repository. This makes it easier for different analyses to be carried out with the same set of data. Historical data helps inform trends in livestock marketing that could help greatly improve policy and decision making in the sub-sector. Data digitisation involves using a mobile application to scan hardcopy data sheets into a smartphone or tablet. Once scanned, a short survey is created to extract this information from the scanned copies to an excel spreadsheet that can then be converted to (Comma-separated values) CSV format for data analysis. CSV is a format that allows data to be read into data manipulation software like SPSS or STATA.

A team of ILRI and SDL personnel set out to digitise the data sheets in the five counties. Since softcopies of data sheets are often unavailable, this is a first step in the process of digitising data. One of the biggest challenges in this process was having to manually hold a tablet over the data sheet to scan it. In instances where there are multiple data sheets to be scanned, the person scanning often gets tired. The mobile application being used to scan the sheets may also slow down and eventually stop. If at this point the data sheets have not been saved, they are lost and have to be rescanned. Also, the data sheets are hand written, making it difficult to decipher what the writing is in case of a blurry image. When sheets are scanned in haste, the scanning app tends to record blurred images. Therefore, it becomes a painstaking exercise which requires a lot of patience on the side of the one scanning.

Scanning of data sheets in Isiolo

Survey form

The survey form is designed to extract all information in the data sheets to make it easier to read into data analysis software. The goal of the exercise is to be as accurate as possible in retrieving data to minimize introducing errors and distorting the data.

Sample survey form

To ensure we retain the quality of data while digitising, a team of data entry clerks were taken through training before they embarked on digitisation. The process requires going through every data sheet to accurately capture data. There are at least 15 entries on every data sheet. One entry represents one observation in the data analysis software; thus, care has to be taken not to repeat figures.

In-depth discussions with supervisors and data monitors in the five counties revealed there is more data available than is captured in the NLMIS. Due to lags in data collection as a result of irregular funding, other agencies collected data at different times over the years. Data captured in NLMIS only includes data that was collected by SDL. As noted earlier, the bulk of data collected does not make it to the server due to several factors that have been highlighted, among them technical glitches. Data monitors are keeping these hardcopy data sheets, that are a treasure trove of livestock market information. These will potentially be lost, without a concerted effort to salvage them by digitising. If we factor in data that was collected by partner agencies, we realise there is need to push this effort further to ensure we have all data that is out there in one repository. There is need for continuous collection of data that is not in the custody of SDL to ensure we have a robust database. At the same time, we need to figure out how to make data collection sustainable so that it does not depend on donor funds that have an end date. Creating a “Google” of livestock market information that creates greater access to information by pastoralists and other users is the end goal. We believe information access will lead to better decision making in the process of buying and selling livestock. This will eventually improve livelihoods and wellbeing of pastoral communities that have been lagging in development for a long time.

Show More

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *


Adblock Detected

Please consider supporting us by disabling your ad blocker