Scaling up successfully: Lessons from Kenya’s Tusome national literacy program

Kenya is a lower-middle-income country with a per capita annual gross domestic product of US$1455 (World Bank 2017a) and a population estimated at 48 million (World Bank 2017b). The 2010 Constitution transitioned Kenya from a centralized system of service delivery to a system decentralized at the county level to increase accountability (Republic of Kenya 2010). The education sector was the largest social sector that did not decentralize financial control, making investigating scaling education a more complex exercise than in other sectors in Kenya.

Low learning outcomes is an area of increasing concern and interest in Kenya. An Uwezo household survey assessment (Uwezo 2016) showed no national-level changes in literacy outcomes from 2010 to 2014, despite the implementation of more than 25 interventions across Kenya in 2014 (Kibukho, personal communication, October 15, 2014). In addition, data from Early Grade Mathematics and Reading Assessments (EGMA and EGRA respectively)Footnote 2 showed that 40% of pupils could not read or comprehend what they read by the end of Class 2 (Piper and Mugenda 2012) and that numeracy outcomes in both conceptual and procedural tasks (Piper et al. 2016c) were lower than Ministry of Education-established performance benchmarks (Reubens 2009). In 2012, on average, only 5% of Class 2 (grade 2) children reached the benchmark fluency rates for comprehension in Kiswahili, and 7% did so in English (Piper and Mugenda 2012).

Kenya has long been a location for rigorous studies examining the impact of educational programs that attempt to work at scale (Bold et al. 2013; Kremer et al. 2009). Kenya is the country in sub-Saharan Africa with the highest number of studies cited in the Abdul Latif Jameel Poverty Action Lab’s (J-PAL’s) database of educational interventions (https://www.povertyactionlab.org/evaluations). In fact, Kenya trails only India and the United States in the number of evaluations in J-PAL’s database.

Kenya’s status as a preferred research site for pilot programs, however, has not protected it from also being the setting for failed large-scale initiatives, including basic education programs of various designs. The implementation of the 2002 curriculum struggled because of its adherence to relatively outdated instructional routines (Kenya Institute of Curriculum Development, KICD 2017). School-based teacher professional development programs struggled because of a perception that the key resource teachers were not chosen by merit (Chege 2011). The government’s system of instructional support through Teachers’ Advisory Centre (TAC) tutors suffered from the burden of administrative tasks, and as a result, by the beginning of Tusome’s implementation in 2015, very few classrooms were being supported by this professional cadre, outside of donor-funded programs (Zuilkowski and Piper 2017). The Strengthening of Primary Education program for head teachers to provide classroom support failed to develop incentives for this instructional leadership function (Cambridge Education Consultants 1998). A large-scale literacy and numeracy intervention in Western Kenya struggled to have teachers consistently implement the program, leading to no impact or negative results (Lucas et al. 2014). An instructional improvement program called Literacy Boost had an impact when managed by a nongovernmental organization but none when managed by the Ministry of Education (Bold et al. 2013). Clearly, Kenya is a context open to research and experimentation, but success at scale has remained elusive.