<?xml version="1.0" encoding="UTF-8"?>
<feed xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns="http://www.w3.org/2005/Atom">
<title>2020 - 2029</title>
<link href="http://hdl.handle.net/11090/965" rel="alternate"/>
<subtitle>Working Papers</subtitle>
<id>http://hdl.handle.net/11090/965</id>
<updated>2026-11-22T00:53:05Z</updated>
<dc:date>2026-11-22T00:53:05Z</dc:date>
<entry>
<title>Gender Disparities in Leisure Time Consumption: Evidence from a Time Use Survey in Kenya</title>
<link href="http://hdl.handle.net/11090/1067" rel="alternate"/>
<author>
<name>Kamundia, Susan</name>
</author>
<author>
<name>Oleche, Martine Odhiambo</name>
</author>
<author>
<name>Manda, Damiano Kulundu</name>
</author>
<author>
<name>Wambugu, Anthony</name>
</author>
<author>
<name>Mwabu, Germano</name>
</author>
<id>http://hdl.handle.net/11090/1067</id>
<updated>2026-11-20T07:39:34Z</updated>
<published>2026-11-01T00:00:00Z</published>
<summary type="text">Gender Disparities in Leisure Time Consumption: Evidence from a Time Use Survey in Kenya
Kamundia, Susan; Oleche, Martine Odhiambo; Manda, Damiano Kulundu; Wambugu, Anthony; Mwabu, Germano
This paper examines gender disparities in the consumption of leisure time, an understudied area of research in Kenya and elsewhere in Africa. The study uses nationally representative datasets from the 2021 Kenya Time Use and the Kenya Continuous Household Surveys. Leisure time is defined as the residual time in a day after accounting for time spent in employment, production for own use, domestic work, care work, volunteer work, learning, self-care and maintenance. First, the study analyses the individual and household factors that determine the amount of leisure time allocated by individuals. Together, descriptive and ordinary least squares results reveal a substantial gender gap in leisure time, with women having significantly less leisure time, even after controlling for covariates of interest, such as age and education. Second, the study disaggregates the gender gap in leisure time using the Oaxaca–Blinder approach, augmented with recentred influence functions. The results show that the disparity in leisure time consumption by gender is mainly driven by structural factors rather than by differences in endowments. Differences in returns to education and labour force participation narrow this gap. However, the presence of young children in the household exacerbates the gender gap in leisure time consumption in favour of men, highlighting the negative role played by entrenched gender norms in widening gender-based inequalities in leisure time allocation across welfare enhancing activities. A Fairlie decomposition of leisure activity preferences shows a general preference for cultural over social leisure, although women are more likely than men to prefer social leisure. Observable characteristics, chiefly education and household size, explain about 47% of this gap, with the remainder unexplained. Overall, the decomposition evidence shows large, gender-based inequalities in the use of leisure time in Kenya.
JEL codes: J16, J22, D13, D63
</summary>
<dc:date>2026-11-01T00:00:00Z</dc:date>
</entry>
<entry>
<title>Developing a local measure of inequality</title>
<link href="http://hdl.handle.net/11090/1066" rel="alternate"/>
<author>
<name>Jacobs, Chandré</name>
</author>
<id>http://hdl.handle.net/11090/1066</id>
<updated>2026-11-13T09:00:56Z</updated>
<published>2026-11-01T00:00:00Z</published>
<summary type="text">Developing a local measure of inequality
Jacobs, Chandré
In this paper, I develop a measure that aims to capture an individual’s awareness of inequality based on where they live and how they might move around in society and call it the Local Inequality Index. This measure is local in two senses of the word. First, it measures inequality from the perspective of an individual rather than giving a summary statistic for the population. Second, it captures the inequality that exists in and around the place in which an individual lives. As such, the Local Inequality Index may also be thought of as a spatial measure of economic inequality linked to the location in which an individual resides. I also show how a global measure of inequality (similar to the Gini coefficient) may be computed by aggregating inequality from the perspectives of all individuals in the population and develop the Local Population Inequality Index. Measuring inequality in a way that encapsulates the experiences of individual members of society is crucial if we are to better understand how inequality affects human behaviour and social outcomes.
This is a joint SALDRU/ACEIR Working Paper
</summary>
<dc:date>2026-11-01T00:00:00Z</dc:date>
</entry>
<entry>
<title>Undocumented imputation and implausible households in the 2022 South African census</title>
<link href="http://hdl.handle.net/11090/1065" rel="alternate"/>
<author>
<name>Budlender, Joshua</name>
</author>
<author>
<name>Thornton, Amy</name>
</author>
<id>http://hdl.handle.net/11090/1065</id>
<updated>2026-10-29T15:09:00Z</updated>
<published>2026-10-01T00:00:00Z</published>
<summary type="text">Undocumented imputation and implausible households in the 2022 South African census
Budlender, Joshua; Thornton, Amy
We identify major data quality problems in South Africa’s 2022 Census data, which suggest substantial undocumented imputation concentrated in 20% of households. Using the public 10% microdata sample, we first show that the distribution of person weights in the 10% sample is unusual and extreme, while the household weight distribution is more typical. We then show that the most extreme person weights (vanishingly small weights of less than 1) are all found in a subset of households which share peculiar item non-response patterns. These households have anomalous and highly implausible characteristics, both for the household as a whole and for the individuals in these households. Close inspection of these households suggests that these implausible characteristics are due to rudimentary and undocumented imputation. The effects of this imputation are substantial and are reflected in official Census statistics as well as the 10% sample: for example, the reported increase in the prevalence of multiracial households from 1.7% in Census 2011 to 11.4% in Census 2022 is almost fully explained by implausible imputation in the subset of households we identify.  We conclude that these households are likely wholly imputed, speculate as to what could have motivated such a procedure, and discuss the implications for typical uses of the Census data. We end by suggesting a constructive path forward.
</summary>
<dc:date>2026-10-01T00:00:00Z</dc:date>
</entry>
<entry>
<title>The mental health dividend of secondary education for Ghanaian women: Evidence from the Ghana Socioeconomic Panel Survey</title>
<link href="http://hdl.handle.net/11090/1064" rel="alternate"/>
<author>
<name>Adzah, Betty Baabo</name>
</author>
<author>
<name>Branson, Nicola</name>
</author>
<author>
<name>Whitelaw, Emma</name>
</author>
<id>http://hdl.handle.net/11090/1064</id>
<updated>2026-09-10T13:21:50Z</updated>
<published>2026-09-01T00:00:00Z</published>
<summary type="text">The mental health dividend of secondary education for Ghanaian women: Evidence from the Ghana Socioeconomic Panel Survey
Adzah, Betty Baabo; Branson, Nicola; Whitelaw, Emma
Across sub-Saharan Africa, widening access to and completion of secondary education – especially for women – has become a central development priority. More schooling is associated with improved psychological wellbeing, but whether completing secondary school in particular is linked to better mental health among women in sub-Saharan Africa is not well understood. Using four waves of the Ghana Socioeconomic Panel Survey (2009/10–2022/23), we estimate the relationship between lower- and upper-secondary credentials and women’s Kessler-10 (K10) distress scores, over and above a linear relationship with years of education. We find that an upper-secondary credential is associated with lower distress net of years of education: about 1.19 K10 points, or a 7.8-percentage-point lower probability of being at risk (K10 ≥ 20). This association is not concentrated among the tertiary-educated in that group. A lower-secondary credential is also associated with a 4.3-percentage-point lower probability of being at risk, net of years of education. The credential gradient is attenuated, but remains, after adjustment for observed economic security, household circumstances, physical health and care burden; economic security produces the largest attenuation.The panel allows us to track the same women between waves. Among women not at risk at one interview, secondary credentials are associated with a lower probability of being at risk at the next; among women already at risk, there is little evidence of a credential gradient in recovery. However, educational attainment is not randomly assigned and the current-circumstance measures are observed contemporaneously with distress, so the estimates are conditional associations and the sequential adjustment is an accounting exercise rather than a causal mediation analysis. The findings nevertheless point to psychological wellbeing as a potentially important non-market return associated with secondary education.
</summary>
<dc:date>2026-09-01T00:00:00Z</dc:date>
</entry>
</feed>
