Conflict and student learning in East Africa

By Charles Gale, Research Associate, Education Policy and Data Center

Overview

The connection between violent conflict and educational access was the subject of a previous post.  In this post we look at the relationship between conflict and learning levels. As children and youth have gained access to schooling at a rate unprecedented in history, attention has shifted to outcomes of the education process, such as literacy and employability. In most research about the relationship between violent conflict and education however, the emphasis has stayed on measures of access and participation. This post presents data from recent sources to speculate about the relationship between conflict and literacy levels in the context of two East African countries that have experienced varying levels of conflict in recent years, Kenya and Uganda. While a thorough discussion of the country contexts and events surrounding the outbreak of conflict are beyond its scope (see below for some useful resources), it does provide some evidence to guide future avenues for research.

Advancements in data availability

Measuring learning levels has typically been the domain of large-scale national and international organizations, and has been conducted at the school level. In recent years, proliferation in the number of so-called “citizens-led” and early grade reading/numeracy assessments have shed new light on learning levels as an outcome of education systems, and has put pressure on a number of governments to start thinking about how to improve the quality of education. There has also been an increase in the use of geo-referenced data, which has proven particularly useful for those researching conflict as it provides data disaggregated to a lower geographic level than was afforded in the past. This post uses data from the most recent rounds of the UWEZO assessment in Uganda and Kenya, as well as data from the Armed Conflict Location and Events Dataset, to investigate in more detail the relationship between violent conflict and literacy.

Conflict and learning in context: Uganda and Kenya

Figure 1 provides some larger context as to the scale of conflict in Kenya and Uganda in recent years. It should be noted that while both countries have seen episodes of conflict in recent years, it would be controversial to label either as a “conflict” country. Uganda had a long-running domestic insurgency involving the Lord’s Resistance Army which resulted in a large number of Internally Displaced Persons and severely disrupted the North of the country over a 20 year period. In 2008, Kenya saw outbursts of post-election violence in a closely contested presidential race. Both countries have been relatively stable in recent years, with small-scale skirmishes over land or those that have broken out along ethnic or religious lines. As can be seen, Kenya has experienced relatively consistent low levels of violence since the election there, while the data above capture only the tail-end of events related to the LRA conflict in Uganda. 

The indicator we calculate to measure literacy in this post is the percentage of pupils who are able to read a paragraph of 2nd-grade level text in a non-local (English or Kiswahili) language[1]. Challenges in measuring literacy have been documented extensively elsewhere, such as in this EPDC policy brief that uses Uganda as a case example, and must be acknowledged. Particularly in regards to the curricula followed in both countries, the early grades of primary are a time where instruction in a dominant language such as English or Swahili may be done unevenly and sporadically. For example in Uganda, the national curriculum stipulates that students in primary 1 through primary 3 should be taught in a local language, although implementation is uneven between urban and rural areas. As we can see above, differences in primary grade language competencies are quite apparent depending on the country and the language. There is a much wider range of competencies in English by district in Uganda, compared to counties in Kenya. Further, roughly half of counties in Kenya have higher English competencies than all districts in Uganda. Swahili is included for Kenya to illustrate an important point—competencies may be very different depending on the language of assessment. For example, it is possible that the distribution of scores for local languages would look quite different from what we see above.

Another way of looking at the relationship between conflict and literacy is by identifying regions with historically high levels of conflict and comparing to current learning levels. We choose the years 2000-2010 to aggregate our conflict data, and the year 2012 as our literacy year. The years 2000-2010 encompass “heightened” conflict periods for both countries, although the conflict in Uganda even during these latter years was considerably more intense than that in Kenya. Figure 5 presents learning outcomes for “high conflict” subnational regions, those with 50 or more conflict fatalities over the period. Children whose learning levels were captured in the UWEZO survey could have been exposed to conflict directly or indirectly at some point from birth to the early years of childhood, a critical time for the development of psycho-social skills. The trend line for Kenya has one stark outlier, Turkana, yet even when excluding it (hover over the point with your mouse and click “exclude”), you can see the trend is still downwards. For Uganda the trend is also downward, as the highest-fatality districts (Gulu, Kitgum and Lira) also have relatively low learning levels. But there are also a few high-fatality districts that have relatively high learning levels, such as in Kunungu, Kaabong and Kotido. 

Conclusion

While caution should be exercised in drawing any kind of definitive conclusions from the presentation of the data here, it can be inferred that in Kenya and Uganda, higher conflict subnational areas tend to have lower educational outcomes. Reasons for this, however, are likely numerous and the ways in which conflict interacts with other factors to result in high or low learning levels is as of yet unexplored in empirical research. Due to data limitations it is likely that things will stay this way, although more timely and disaggregated data mean that the ability to investigate the relationship is there in a way that it wasn’t in the past. Below are some additional resources.

·         The 2011 Global Monitoring Report remains the largest and most comprehensive resource on research into the relationship of conflict and education.

·         While there is a wealth of academic research on the LRA conflict in Uganda, a good starting point may be the Advisory Consortium on Conflict Sensitivity’s Northern Uganda Conflict Analysis.

·         The Global Coalition to Protect Education from Attack has a country profile page for Kenya.

·         The USAID-commissioned Literacy and Education in Conflict and Crisis-Affected Contexts provides a framework for those looking to promote literacy education in conflict environments, and the Inter-Agency Network for Education in Emergencies hosts a website with numerous publications for researchers and practitioners.