Updates on private returns to education in Uganda: evidence from universal primary education policy

ABSTRACT

This study updates private returns to education in Uganda using consumption measures as an outcome variable, focusing on obtaining estimates using instrumental variables based on the introduction of the universal primary education policy. Unlike common findings from developed countries, the evidence from this study for a low-income country suggests that returns to education are smaller for liquidity-constrained individuals than for average individuals. The finding also suggests that smaller returns are observed for self-employed farmers than self-employed workers in non-agricultural sectors. The trend, showing smaller returns for liquidity-constrained individuals, is particularly obvious among self-employed farmers.

KEYWORDS:

• Returns to education
• instrumental variables
• universal primary education policy
• Uganda

Acknowledgment

We are grateful to the Uganda Bureau of Statistics for having granted us permission to use the UNHS 2016/17 data sets.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 There was also a lack of commitment to expanding free primary education under Uganda’s leadership during 1960s after independence (Oketch and Rolleston 2007).

2 The UPE policy was proposed by Ugandans before the EFA Movement was launched in 1990. The universalization of primary education was included as one of the key recommendations of the Education Policy Review Commission in 1989. The commissions were appointed by the post-conflict government and were led by a professor from Uganda. A gradual approach to the UPE was recommended at this point (Ssekamwa 1997).

3 Survival rate to grade seven, which is the last grade of primary education, was 32% in 2016 (MoES 2017). Museveni promised to launch the Universal Secondary Education (USE) policy during his 2006 presidential election, and it was introduced in 2007. MoES (2017) reports only a modest increase in net enrollment rate, namely from 15% in 2005 to 24% in 2009.

4 Categories of skill levels for the occupation dummies were created based on the latest International Standard Classification of Occupations (ILO [International Labour Organization] 2023).

5 In the UNHS 2019/20, 15,786 nationally representative households were selected using a two-stage stratified sampling design. According to UBOS (2021), the COVID-19 lockdown was initiated abruptly after the field visits to 6,925 selected households were completed, and the remaining 8,861 selected households were visited during the COVID-19 pandemic. Considering the background of how the survey was conducted, the subsample of data collected before the pandemic cannot be considered nationally representative of the UNHS 2019/20. Furthermore, although the Uganda National Panel Survey (UNPS) collected more recent data than in 2017, this study did not use these because of the unavailability of parental education variables.

6 In UNHS 2016/17, a set rate before taxes and other deductions were removed was reported. For individuals who reported hourly pay, wages were first multiplied by the reported total hours they usually worked per week. We multiplied this value by 52/12. For individuals who reported daily pay, wages were first multiplied by the total number of days they worked per week. We multiplied this value by 52/12. For individuals who reported weekly pay, we multiplied their earnings by 52/12, whereas for individuals who reported being paid every two weeks or twice monthly, the wage was multiplied by two to obtain monthly pay.

7 The number of workers in each household was calculated by adding the number of household members who worked as employees, employers, own-account workers, or family workers and those who looked for work or tried to start their own business for a 30-day period prior to the interview.

Additional information

Funding

This work was supported by the Japan Society for the Promotion of Science under KAKENHI Grant Numbers JP21K13540, JP22H00079.