The negative health spillover effects of universal primary education policy: Ethnographic evidence from Uganda

ABSTRACT Scholars of global health have embraced universal education as a structural intervention to prevent HIV. Yet the costs of school, including fees and other ancillary costs, create an economic burden for students and their families, indicating both the challenge of realising the potential of education for preventing HIV and the ways in which the desire for education may produce vulnerabilities to HIV for those struggling to afford it. To explore this paradox, this article draws from collaborative, team-based ethnographic research conducted from June to August 2019 in the Rakai district of Uganda. Respondents reported that education is the most significant cost burden faced by Ugandan families, sometimes amounting to as much as 66% of yearly household budgets per student. Respondents further understood paying for children’s schooling as both a legal requirement and a valued social goal, and they pointed to men’s labour migrations to high HIV-prevalence communities and women’s participation in sex work as strategies to achieve that. Building from regional evidence showing young East African women participate in transactional, intergenerational sex to secure school fees for themselves, our findings point to the negative health spillover effects of Uganda’s universal schooling policies for the whole family.

Scholars of global public health have widely embraced universal education as a structural intervention for improving individual and community health, including advocating for schooling as a protective factor against Human Immunodeficiency Virus (HIV) (Behrman, 2015;Neve et al., 2015).In Uganda, for example, rising school enrolment associated with the country's Universal Primary Education (UPE) policy has been linked to the following: lower HIV prevalence; delays in adolescent initiation of sex; increases in consistent condom use among young men and women; the increased use of contraception among young women; and reductions in sexual partners and sexual concurrency among young men (Santelli, Edelstein, et al., 2015;Santelli, Mathur, et al., 2015).
Yet the costs of school, which include fees, uniforms, transportation, meals, textbooks, and other ancillary expenses, create an economic burden for students and their families, indicating the challenge of realising the potential of universal education for preventing HIV.Moreover, both desires and legal requirements for primary education may produce vulnerabilities to HIV for those struggling to afford it.Robust evidence from Southern and East Africa indicates that adolescent girls and young women participate in transactional sex with older men to raise money for tuition, exchanging 'sex for fees' in intergenerational relationships that significantly increase their vulnerability to HIV (Bantebya et al., 2014;Dana et al., 2019;Gausman et al., 2019;Kamndaya et al., 2016;Karim & Baxter, 2019;Leclerc-Madlala, 2008;Mojola, 2014;Prince, 2012).Building on these findings, our data show that the economic burden of schooling impacts not just female students but the entire family, including male students, young fathers, mothers, and older siblings, who participate in various forms of intergenerational and transactional sex to raise money to pay for primary school for family members.Our analyses suggest that the high costs of universal education may be a social determinant of HIV acquisition, and that they certainly cause emotional suffering for families.We describe these complex consequences for families as the unintended 'health spillover effects' of an otherwise well-intentioned and well-respected social policy.

Universal primary education in Uganda
Primary education is both a social norm and legal requirement in Uganda, which implemented Universal Primary Education (UPE) in 1997.The UPE policy mandated that every family send at least four children to school and eliminated tuition fees for public primary schools across the country (Bategeka & Okurut, 2006).Universal Secondary Education (USE) followed in 2007, which identified one school per sub-county to be designated as a free government school (Huylebroeck & Titeca, 2015;Molyneaux, 2011).The Ugandan government intended to use universal education policies to increase the social mobility of low-income families, and international observers have lauded UPE as an 'African success story' for providing a seeming ticket out of poverty for Ugandan families (Huylebroeck & Titeca, 2015;Molyneaux, 2011).
Both policies succeeded in raising rates of enrolment and retention in government schools, but they did not make schools entirely free.Families enrolling children in public schools bear the burden of the ancillary costs of schooling as they must pay for uniforms, textbooks, school lunches, and exam registration fees.Moreover, because UPE made primary education a legal requirement, caregivers who forego paying for school (and ancillary costs) are subject to legal punishment (Mwesigwa, 2015).Elder Ugandans remember scenes of police beating parents who refused to send children to school broadcast on national television at the time UPE was implemented, which they interpreted as a warning to those questioning the policy.
UPE and USE also strained publicly funded schools by creating larger class sizes, which worsened academic performance among attendees (Grogan, 2006;Nishimura et al., 2008).Government schools' reputation for providing poor quality education motivates many Ugandan families to pay large sums of money to send their children to private schools, despite the nominal availability of low-cost public schooling (Why Private Outperform UPE Schools, 2021).As one respondent in our study described, 'If you send your child to a government school, it's like you don't care about your child'.According to Uganda's Ministry of Education and Sports, faith-based organisations oversee most educational institutions in Uganda, including 75% of primary schools and 56% of secondary schools.
Uganda's education system follows the British model, institutionalised by missionaries during the country's period as a colonial protectorate .Pre-primary school or 'baby class' is followed by seven years of primary school (P1-P7), four years of lower secondary school (S1-S4, the Ordinary-Levels), and two years of upper secondary school (S5-S6, the Advanced-Levels).Some students attend boarding schools as early as P1, while others begin boarding when they enter secondary school.Ugandan law requires students to remain in school through S4, or until at least the age of sixteen years, but many leave school before that age and others spend significant periods of time away before returning to their education.Some young people may not enter early primary school until their late teenage years, and some secondary schools have students in their early twenties finishing the Ordinary-Levels.School fees are paid once per term, three times per year.

Unemployment, transactional sex, and family-level health effects
Whether sending children to privately or publicly funded schools, Ugandan families struggle to find the money required to educate their children because waged labour is increasingly unavailable, especially to young people.As of 2017, 47% of Ugandan youth were unemployed, a figure that includes unskilled labourers as well as university graduates.Even those who find work are significantly underemployed (Asiimwe, 2022).
In a national context of widespread unemployment, many young people seek financial support through varied forms of transactional sex and romance, intimate-economic arrangements long documented throughout Southern and East Africa (Cole & Thomas, 2009;Stoebenau et al., 2016;Swidler & Watkins, 2007;White & Kamya, 2021).These arrangements include sex sold for a price negotiated beforehand, referred to in Ugandan English as 'commercial sex work', and longer-term relationships with older women and men, or sugar mommies and daddies (Parikh, 2016;Schmidt-Sane, 2021).Coupled with the unavailability of other opportunities for generating income, these and other diverse forms of transactional sex create pathways to HIV vulnerability by increasing the likelihood of intergenerational sex, the number of concurrent partners, and the number of sexual interactions (Higgins et al., 2014;Kirby, 2008;Santelli, Edelstein, et al., 2015;Schuyler et al., 2017).Families struggling to afford school fees also face significant psychoemotional distress.Understanding these complex health consequences of an internationally respected social policy requires nuanced attention to the everyday lives of Ugandans and to locally salient meanings of health and well-being.

Methodology
This article is based on team-based ethnographic fieldwork conducted by researchers at Columbia University and the Rakai Health Sciences Program (RHSP) between June and August 2019.That fieldwork was part of a larger interdisciplinary study investigating how the social transition to adulthood shapes HIV vulnerabilities among adolescents and young adults in the rural Rakai district of Uganda, and whether and how transitioning to adulthood in different types of Rakai communities influences those vulnerabilities.Following Urie Bronfenbrenner's (2000) ecological systems theory, the ethnographic study sought to identify and compare the micro-, meso-, and macro-level systems shaping individual transitions to adulthood and to identify linkages between these nested systems and HIV vulnerability.The research presented in this article was approved by the Institutional Review Boards at Columbia University (AAAR5428), the Uganda National Council of Science and Technology, and the Uganda Virus Research Institute.All study participants provided written consent, documentary evidence of which is held by RHSP and can be supplied if requested.

Study setting and site selection
The Rakai district of southwestern Uganda borders Tanzania in the south and Lake Victoria, Africa's largest lake, in the east.As of 2014, its population was approximately 516,000 (Uganda Bureau of Statistics, 2017).Once one of the world's regions most devasted by HIV/AIDS, Rakai now includes communities many researchers refer to as 'HIV hotspots' (Ratmann et al., 2020).Rakai is primarily rural, with larger, urbanised towns dispersed throughout the district and fishing communities located along the lakeshores.
Study sites were selected using demographic data collected by the longitudinal Rakai Community Cohort Study (RCCS).Since 1994, the RCCS has collected population and sexual and reproductive health data from an open cohort of residents aged 15-49 years from 50 communities in Rakai.Six communities were selected as representative of the three different key types of communities in the area: fishing communities, agrarian communities, and urbanised trading centres.Two demographically similar communities of each type were included in the study to identify patterns across community types.
Descriptive statistics collected by the RCCS illustrating educational attainment, community-level patterns of HIV prevalence, and HIV risk factors for adolescents and young adults contextualise our ethnographic findings.For example, as the RCCS data presented in Table 1 indicate, nearly everyone surveyed in the region has attended at least some primary school (95.1%), pointing to both the success of UPE in raising school enrolments as well as the ubiquity of the economic burden of schooling.RCCS data further indicate that Rakai's fishing communities, when compared with nearby agrarian and trading communities, are places where significantly elevated rates of HIV prevalence converge with the greater presence of key HIV risk factors for young people, including sexual activity among 15-19-year-olds, two or more partners in the past year, inconsistent condom use, and high rates of reported transactional sex, defined as an exchange of money, favours, or gifts for sex with a partner (Chang et al., 2016).These data suggest that fishing communities present unique social-epidemiological environments which shape HIV vulnerabilities for young people migrating to the lakeshores seeking work (Kreniske et al., 2019).

Ethnographic methods and data analysis: Inductive, iterative, collaborative
Ethnography is rooted in grounded theory, a methodology that combines data collection and analysis to allow research questions and instruments to be adapted as data are collected (Strauss & Corbin, 1997).Employing this iterative approach, the research team, which included eight qualitative researchers and native Luganda speakers from RHSP and three researchers from Columbia, met weekly over the three-month data collection period to compare findings, establish patterns in the data, and reformulate research questions.Importantly, rooted our research design in grounded theory made room for data collectors to suggest new lines of inquiry, which led to the present article's focus on the economic burden of schooling.The first phase of the ethnographic study was designed to ascertain community perceptions of young people's health vulnerabilities as well as local understandings of adolescence, youth, and adulthood.For this phase, the research team, directed by Neema Nakyanjo and including Rosette Nakubulwa, utilised multiple qualitative methods, including six community mappings, participant observations, 43 key informant interviews with community leaders, 24 focus group discussions with older and younger men and women, and 26 in-depth interviews with young people and their romantic partners.Data collectors worked in pairs of two, one female and one male data collector per site.Dr. Erin V. Moore, who has been working in Uganda and studying Luganda since 2009, and Bridget Morse-Karzen, then an MPH student at Columbia, accompanied teams to study sites to collect fieldnotes.
To initiate the study, data collectors worked with a community contact already established by RHSP to conduct a topographical survey of each community, including its key points of business, schools, religious institutions, biomedical and allopathic healthcare providers, places of leisure, and residential dwellings, and to identify key spaces where young people live, work, and socialise.Data collectors drew maps reconstructing the survey, which were used to guide ethnographic participant observations and further sampling for study participants.Each data collection team visited the same community 3-4 times per week over the course of six weeks.During these repeated visits, data collectors took detailed notes on their observations of the social interactions unfolding around them.
Working from the community mappings and with community partners, data collectors identified and interviewed recognised leaders of business, health, religion, and education in each community (Table 2).Community leaders were asked to identify salient features of adulthood for young men and women and to reflect on the key issues young people in their community face.Community leaders then helped data collectors identify older and younger men and women to participate in focus group discussions (FGDs) on what it means to transition to adulthood for young men and women, how that has changed between generations, and what key issues contemporary young people face as they become adults.From the FGDs with younger men and women, young people and their current romantic partners were invited to participate in-depth interviews.Then, in separate, private interviews, data collectors asked interviewees and their romantic partners to describe their life histories, what motivated their migrations to their community of residence, how their romantic partnership had formed, whether each partner considered themselves adults, what constitutes adulthood, and what important issues they felt they and other young people in their community face.
Working in pairs, one data collector conducted the interview or FGD while the other recorded observations in fieldnotes.Data collectors tape-recorded FGDs and interviews, which they translated from Luganda to English the following day.Data collectors submitted fieldnotes, translated interview and FGD transcripts, and one-page memos analyzing the observation, interview, or FGD in relation to the themes of the broader study to a designated data manager, who compiled, organised, and ensured the security of the data.Data collectors also submitted photographs documenting the built environment in each community.
During the data collection period, Dr. Moore reviewed and coded all data compiled each week and developed a dynamic codebook from themes recurring in the data.Once per week, data collectors, Dr. Moore, and Ms. Morse-Karzen met in person or over Zoom to discuss the prior week's research activities, reflect on the research process, report key findings, and identify 'keywords', which were added to the codebook.These discussions lent themselves organically to drawing comparisons across sites, which led the research team to collectively identify areas to probe or new questions to pursue the following week.
This iterative analysis process quickly surfaced the centrality of the economic burden of schooling in the everyday lives of the study population, with community members and young people alike linking the costs of schooling to HIV vulnerability.Importantly, school fees were not among initial research topics.Rather, the issue of school fees emerged unprompted during conversations with study participants about health, adulthood, and everyday life.For example, when asked what the most important health problem young people faced, respondents in our study overwhelmingly reported 'lack of income', which led the research team to further questions about the everyday expenditures required of young families and how they managed to afford them.
Once sensitised to the burden of school fees in Rakai community members' everyday lives, the research team further investigated the actual costs of schooling for Rakai families.National and district-level data on the costs of primary education for Ugandan families are difficult to obtain, especially because ancillary expenses vary so much by school, so the team conducted a community school survey to inquire about required fees and ancillary expenses at the privately-and government-funded primary and secondary schools in each study site.In most sites, the team interviewed the school directors from every school in the immediate area; in urban communities, the team selected the most attended schools as identified by community leaders.School directors were asked what fees or in-kind goods (including items such as kilos of grains, cooking utensils, and toilet In-depth Interviews (n = 26) -Young people, 17-28, and their romantic partners -Young men ( 14) -Young women (12) From the FGDs with younger men and women, young people with current romantic partners were asked to participate in-depth interviews.In separate, private interviews, data collectors asked FGD participants and their romantic partners to describe their life histories, what motivated their migrations to their community of residence, how their romantic partnerships had formed, whether each partner considered themselves adults, what constituted adulthood, and what important issues they felt they and other young people in their community face.
paper) were required for one student per term.In a second phase of analysis, these data on the real costs of schooling faced by families were triangulated with data collected during interviews and FGDs that reflected community members' perceptions of the impacts of the cost burden of education on family health.

Global health research collaboration
Social scientists have increasingly interrogated the various and often unequal forms that global health research partnerships take in practice, which frequently proceed with scientific questions formulated in the global North and data collected in the global South (Abimbola, 2019;Kenworthy et al., 2018).Wanting to take a more equitable approach to global health research collaboration, the research team intentionally organised analysis meetings to make space for question formulation by Ugandan scientists who were collecting data.It was through the process of collaborative analysis that the economic burden of schooling emerged as a crucial topic for further investigation, not least because many members of the research team could reflect upon the ongoing stresses of raising school fees in their own lives.That data collectors, university-educated research scientists, and research participants, who were generally of much lower socioeconomic status, shared concerns about school fees suggests that the cost burden of education transcends class difference in Uganda.

Results
Of 69 individuals interviewed in Rakai, half (34) reported that 'school fees', a phrase used to signal both tuition and ancillary costs, were the most significant economic burden faced by Ugandan families, even more significant than housing expenses.Interviews with school administrators revealed that even UPE schools, nominally free for students to attend, require costly expenditures.In Rakai's urban areas, for example, fees associated with government-funded schools can amount to as much as 17% of a household's income per student, a figure calculated based on an average household income of $2 USD/day; private schools fees can amount to as much as 66%.As we describe below, 'finding' school fees, as Ugandans say, is both legally mandated and fundamental to local ideals for masculine adulthood, which motivates men to migrate to seek incomes to pay for their children's schooling.Strategies for finding school fees are gendered, with young men and women pursuing different avenues to secure funds to educate themselves and their family members.

Local conceptions of masculine adulthood
The ability to pay school fees for one's children is central to Rakai residents' conceptions of what it means to be an honourable adult, particularly for men.Older men remarked that finding fees was the mark of the transition to masculine adulthood.As one older man, a farmer, told us, A boy becomes an adult when he leaves his parents' home and builds a house and marries a woman.He then starts his duties of paying school fees for his children, buying food at home and the rest.This makes him a man.
Young men likewise aspire to this ideal of masculinity, to 'work hard to prepare for the children', as they describe, even as they are increasingly excluded from the wage-earning work through which they might do so.One young man, a welding apprentice, explained, 'A good man is one who looks after his family in a way of giving them needs, like school fees'.Because a man's successful transition to adulthood is demonstrated by his ability to provide for his children's education, many men strive to find ways to afford the costs of school.Some fail, leaving the responsibility to women and other family members.Young women responsible for paying school fees also described 'working for my children' as a key marker of reaching adulthood.

Gendered pathways
The pathways linking young people seeking funds for school fees to HIV vulnerability are complex, manifold, and distinctly gendered.First, for young men in southcentral Uganda, the search for school fees, whether for themselves or family members, sets into motion a series of life events that revolve around seasonal migrations to Lake Victoria, where they find low-wage work in the fishing industry and take up temporary residence in lakeshore communities, where HIV prevalence is disproportionately high, as discussed above.Second, southcentral Ugandan women report participating in dangerous forms of commercial sex work they would otherwise avoid, including sex without condoms and with diminished capacity for refusal, to raise money to pay for their children's education, a finding consistent with studies conducted elsewhere in East Africa (Cooper et al., 2017).
Young men, migration, and intergenerational sex Lake Victoria's fisheries offer seasonal labour opportunities for young men, who migrate to the lakeshores seeking work detangling fishing nets or launching boats to the waters.Community leaders and young men we spoke with overwhelmingly ascribed these labour migrations to men's search for school fees, both for themselves and for family members.Some young men return to school with their wages, but many more leave school altogether to continue earning money.As one seventeen-year-old described, 'It's like a disease … the moment someone starts to make money, you say to yourself, by the time I will sit in class for two hours I will have made 20,000UGX (Ugandan shillings, or USD $5.60)'.Young men report abandoning school despite desires to continue education.As a sixteen-year-old net-detangler asserted, 'What hurts me most is not going to school, because I loved studying … because now you can't even stand for a political office if you're not educated like that'.Young men's migrations to the lakeshores make them vulnerable to HIV because they move into communities where various forms of transactional, intergenerational sex are common and where the prevalence of HIV is disproportionately high, particularly among older women.RCCS data indicate that 51.5% of women ages 25-49 in the fishing communities we studied were HIV-positive (see Table 1 above), and young male migrants report both buying and trading sex with older women.Because younger men make so little money at the lakeshores, they report and are reported to prefer 'commercial sex work', or sex for a price negotiated ahead of time, to longer-term relationships with women their own age, which require ongoing investments of money and gifts.As one sex worker described, 'Many young men have sex with older women because of money.They state that young boys do not want to date girls of their age because they demand a lot of money from them which they do not have'.Young men also reported preferring sex with older sex workers because they charge a lower price than younger sex workers, and, as described below, are willing to offer condomless sex for higher amounts.
In addition to purchasing sex from older female sex workers, young men also participate in sugar mommy relationships, wherein they receive money, alcohol, meals, and housing from older women who own bars and restaurants.An imam we interviewed put it this way, 'If a young man is poor, he might be deceived by an older woman and sometimes that woman is infected.He might not mind whether he is infected and because of poverty he gets married to that woman'.Community members understand relationships with older women as making young men vulnerable to HIV and thus more likely to transmit HIV to women their age.
Women, motivations for sex work, and condomless sex As part of the broader ethnographic study, data collectors spoke with sex workers in a large town along the Trans-African Highway as well as with younger women in other communities who were enrolled in programmes designed to divert them from entering sex work.Young women yet to have children old enough to send to school spoke generally about the links between school fees, transactional sex, and HIV vulnerability.For example, when asked how many children she intended to have, one young woman responded, I first study the situation, then first have two children.If one is hard to raise even with money, what about the four children.Food is scarce and school fees are going high every day.Uganda is developing, and school fees are like in millions.A cheap school costs 150,000UGX ($39), sometimes a man can lack money like the one who takes alcohol and does not know his responsibility at home.So, if you give birth to four children, you're tempted to go for sex work, selling home assets, getting an HIV-infected man because you want to get money to pay your children's school fees.
Whereas younger women described school fees as creating potential for entering risky transactional sexual relationships, older women already engaged in sex work indicated they unequivocally understand their vulnerability to HIV as linked to the costs of schooling their children.They explicitly linked their motivations for participating in sex work, and moreover their preference for sex without condoms, to their desires to educate their children.As one older woman described, referencing locally established prices for sex with and without condoms, Another thing, I can be having my problem: my child has been chased for school fees.So, you are not going to tell me, 'Aaha, I give you 3,000UGX ($0.78)'.I tell you to give me 10,000UGX ($2.61) because my dream is to pay fees for my child.I have to get it from you.And you must do it fast so that I get another customer.
In this statement, this woman explained that she preferred condomless sex with as many partners as possible in order to quickly raise money for school fees.Another sex worker linked school fees to the inability to refuse men's requests during sex.As she put it, When a man gives you his money, he starts playing with your body, like he can say, 'I bought you with my money, put your leg up like this.Don't say anything, I gave you my money' … if you refuse, he tells you to refund his money.Since he has given you his 10,000UGX ($2.61) or 20,000UGX ($5.20) and you have a problem of school fees, you end up accepting.So, you end up getting infected or infecting him with your STIs.By the way, those men infect us with many diseases.
As these comments indicate, sex workers see the need to pay for children's schooling as a motivation for condomless sex, and they blame school fees both for making them vulnerable to sexually transmitted infections and for limiting their agency during sexual encounters.Importantly, these data should not be taken to indicate that older female sex workers are simply 'vectors of risk' who intentionally, even maliciously seek to spread HIV among young men.To be sure, sex workers regularly contend with discriminatory social narratives that deem their survival-based actions as proof of their deceitfulness and suggest that they are a group that should be feared rather than understood and supported.By contrast, the women we interviewed explicitly tied their participation in sex work to poverty exacerbated by the need to raise money to send their children to school.

Discussion
Our analyses advance the robust literature linking high rates of HIV prevalence among young East African women to the exchange of sex for tuition.This phenomenon is often connected to upwardly economically mobile girls trading sex to pay for secondary school or university (Mojola, 2014).Our findings reveal that the health risks of school fees extend beyond female students to parents and siblings, affecting the whole family.These risks are created by legal and social imperatives to educate one's children in a context where an under-funded school system transfers many costs to individual families.Our findings indicate that the search for school fees creates two potential pathways to increased HIV vulnerability for residents of southcentral Uganda.First, for young men seeking funds for education through seasonal labour migrations, Lake Victoria's fishing communities predispose them to forms of intergenerational and transactional sex linked to increased HIV vulnerability, particularly as half the women living in fishing communities are HIV-positive.Second, for some women, the costs of sending children to school motivate their participation in sex work, including incentivizing condomless sex for higher payment.
In addition to increased HIV vulnerabilities, families face significant psycho-emotional stress when it comes time to find fees each term.During collaborative analyses meetings, the research team reflected upon the stress and anxiety study participants demonstrated when discussing the cost of education.Families are reported to have to choose between paying for life-saving medications to treat HIV and other chronic illnesses and their children's school fees in the months they come due (Nolen, 2023;Whyte, 2014).Other scholars have argued that the inability to pay for children's schooling threatens men's ability to fulfil their masculine responsibilities to provide, which may give rise to intimate partner violence (Murray et al., 2021;Smith, 2020;Yllö & Torres, 2016).By contrast, family-level economic interventions in Uganda have been shown to improve overall family health, lower adolescent risk-taking behaviour, and improve school performance (Cavazos-Rehg et al., 2021;Ssewamala et al., 2010Ssewamala et al., , 2020)).Likewise, South Africa's Child Support Grant, an unconditional cash transfer to support parents for the costs of up to six biological children, has also been proven to improve family mental health and HIV outcomes (Booysen, 2004;L. Cluver et al., 2013;L. D. Cluver et al., 2016;Ohrnberger et al., 2020).In contexts of widespread unemployment where families benefit from financial assistance, schooling offers both a path to opportunity and an economic burden that negatively affects family health.

Conclusions
For scholars of global public health, schooling has long been considered a key structural intervention for improving the health of students and their families.Uganda's UPE policy has been especially lauded as one of the great success stories of universal education, studied by demographers for decades.Yet our ethnographic research investigating how young people and their families are currently affected by UPE suggests this vital social policy may also have significant unintended consequences.These findings have important implications for public health researchers and practitioners.Namely, those seeking to explore correlations between universal schooling policies and health must first understand how these policies unfold in the everyday lives of the people they intend to benefit.How accessible are schools supported by universal schooling policies, and do families understand these schools as desirable for their children?
More broadly, our study shows that hearing what people have to say about the issues that matter the most to them without already presuming to know the significant health issues they face can yield surprising, yet crucial, information for improving public health and well-being.By taking an inductive, ethnographic approach to investigating HIV vulnerabilities among adolescents and young adults in southcentral Uganda, and more specifically by asking open-ended questions to community members about what they understood to be key risks to young people's health, we uncovered complex, family-level health spillover effects associated with Uganda's universal schooling policiespolicies which do not, in practice, result in free education.These policies were designed to pull Ugandans out of poverty, but our data reveal that they end up producing health vulnerabilities, particularly among Uganda's poorest citizens.

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

Table 1 .
Grabowski et al. (2017) and epidemiological characteristics for fishing communities vs. agrarian communities and trading centers in Rakai, 2016-2018.Key risk factor for HIV among adolescents and young adults.^Educationattendedcomes from the survey questions: Have you ever gone to school?To what level?^^Transactional sex comes from the survey question: Were money, gifts or favours exchanged for sex with this partner?(4mostrecent partners) For RCCS research design and procedures, seeGrabowski et al. (2017).

Table 2 .
Description of ethnographic methods and sampling.