Social Sciences & Humanities Open 11 (2025) 101311 Available online 24 January 2025 2590-2911/© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). Regular Article Does free secondary education mitigate the impact of teen births on educational outcomes? Evidence from Ghana Kwasi Gyabaa Tabiri a,* , Muniratu Tahiru a, Yaw Boateng Atakorah c, Jacob Novignon b a Department of Economics Kwame Nkrumah University of Science and Technology, PMB, University Post Of昀椀ce, Kumasi, Ghana b Center for Social Policy Studies University of Ghana, Accra, Ghana c Department of Economics Education Akenten Appiah-Menka University of Skills Training and Entrepreneurial Development, Kumasi, Ghana A R T I C L E I N F O Keywords: Teen childbirth Educational attainment Instrumental variable Ghana Free secondary education Sub-saharan Africa A B S T R A C T Teenage pregnancy and early childbirth have potential implications for human capital development that can transcend the teen mothers to their children. Using data from the latest demographic and health survey data from Ghana, this study set out to understand how teen births affect the educational attainment of the mother, and how a policy that reduces the 昀椀nancial cost of senior high school education could moderate this effect. We used instrumental variables to account for endogeneity of teen births. Our results show that teen birth has a negative impact on the educational attainment of the mother. Furthermore, a program that reduced the 昀椀nancial cost of senior secondary education helped attenuate this negative impact for girls in urban areas but had no such effect for girls in rural areas. Our 昀椀ndings emphasize the need to intensify efforts to provide support systems for teen mothers that will enable their reintegration into school after childbirth. They also point out the importance of targeting such interventions to ensure that they address groups of people with peculiar challenges including relatively poor infrastructure, economic conditions and cultural practices/beliefs that facilitate risky sexual behavior. 1. Introduction Teenage pregnancy continues to pose major challenges to public health, especially across developing regions. It is estimated that as of 2019, about twenty-one million girls aged between 15 and 19 become pregnant annually in developing regions. Of these, it is estimated that about twelve million end up giving birth. (WHO and UNICEF, 2023). Worldwide, adolescent birth rate (ABR)1 has signi昀椀cantly declined over the past two decades, falling from 64.5 births per 1000 adolescent girls in 2000 to 41.3 births per 1000 adolescent girls in 2023 (WHO, 2023). Despite this observed progress, wide disparities exist between regions around the world. For instance, sub-Saharan Africa, with the highest rate globally, has an ABR of 97.9 births per 1000 - more than twice the global average, while the European region has a rate of 13.1 per 1000 (less than half the global average) (WHO, 2023). Several factors have been identi昀椀ed to explain the high rates of teenage pregnancy across developing countries. Available evidence suggests that girls initiate sexual activity at a much younger age (Kassahun, 2019; Kyei-Arthur, Agyekum, & Kyei-Gyam昀椀, 2024). At that stage, there is little knowledge of avoiding unplanned pregnancies and sexually transmitted infections. The temptation to conform to peer pressure also increases the chances of early and unprotected sexual ac- tivity (Gyan, 2013; Chandra-Mouli et al., 2014; Gyesaw & Ankomah, 2013). Teenage pregnancy and early childbirth have potential implications for human capital development that can transcend the young mothers to their children. This is evident in the high rate of school dropouts that result from teenage pregnancy and childbirth (Groves et al., 2022; Sobngwi-Tambekou et al., 2022). The high rate of dropout by teenage mothers suggests they are less likely to participate in the labor market, and if they do, their productivity may be relatively low. Consequently, teenage mothers who drop out of school stand a higher chance of being unemployed in the future due to their relatively low human capital development. In part, the school dropout of teenage mothers after * Corresponding author. E-mail addresses: kgtabiri95@gmail.com (K.G. Tabiri), tmuniratu@gmail.com (M. Tahiru), atakorahboateng@gmail.com (Y.B. Atakorah), jnovignon@ug.edu.gh (J. Novignon). 1 Abbreviations: ABR – Adolescent Birth Rate; DHS – Demographic and Health Surveys; FSHS – Free Senior High School; GDHS – Ghana Demographic and Health Surveys; GES – Ghana Education Service; GSS – Ghana Statistical Service; SHS – Senior High School. Contents lists available at ScienceDirect Social Sciences & Humanities Open journal homepage: www.sciencedirect.com/journal/social-sciences-and-humanities-open https://doi.org/10.1016/j.ssaho.2025.101311 Received 11 November 2024; Received in revised form 15 January 2025; Accepted 20 January 2025 https://orcid.org/0000-0001-8688-9979 https://orcid.org/0000-0001-8688-9979 mailto:kgtabiri95@gmail.com mailto:tmuniratu@gmail.com mailto:atakorahboateng@gmail.com mailto:jnovignon@ug.edu.gh www.sciencedirect.com/science/journal/25902911 https://www.sciencedirect.com/journal/social-sciences-and-humanities-open https://doi.org/10.1016/j.ssaho.2025.101311 https://doi.org/10.1016/j.ssaho.2025.101311 http://creativecommons.org/licenses/by-nc-nd/4.0/ http://creativecommons.org/licenses/by-nc-nd/4.0/ Social Sciences & Humanities Open 11 (2025) 101311 2 pregnancy or childbirth is attributable to stigma they face in their communities and schools (Neal et al., 2020; Yakubu & Salisu, 2018). In Ghana, the ABR among girls aged 15–19 years was estimated to be 64 births per 1000 in 2021 (World Bank, 2024). While this is lower than the African average, it is still much higher than the global average. The 2022 Ghana Demographic and Health Survey (DHS) reports that preg- nancy rate in girls aged 15 to 19 ranges from 6% to 8% in the Greater Accra and Ashanti regions respectively, to 24% and 26% in the North East and Savannah regions respectively. In terms of locality, 8% of girls (aged 15 to 19) living in urban areas and 14% of those living in rural areas had ever had a live birth. Given the potentially negative health and human capital implications of adolescent births, the government of Ghana has over the years developed policy frameworks towards curbing this phenomenon. For instance, in collaboration with the Ministry of Health and the United Nations Population Fund (UNFPA), Ghana’s Ministry of Gender, Children, and Social Protection developed a 昀椀ve-year (2018–2022) strategic plan to address adolescent pregnancy in Ghana.2 The Ghana Education Service (GES) also developed a set of guidelines aimed at preventing pregnancy among schoolgirls and facil- itating re-entry for those who do get pregnant (Ghana Education Service, 2018). The documents and programs recognize both the potential of education to improve the future economic and social outcomes of ado- lescents and young girls in particular, and the adverse effect that early pregnancy can have on this potential. The Free Senior High School (FSHS) policy was introduced in Ghana in 2017 with the aim of removing 昀椀nancial barriers and ensuring equity in the access of quality secondary education.3 While the policy was not targeted speci昀椀cally at promoting the education of adolescent mothers, the reduction of 昀椀nancial constraints can help adolescent mothers continue their education. Indeed, studies have shown that 昀椀nancial constraints associated with childbirth can contribute to school dropout among adolescent mothers (particularly those from low-income house- holds), and that providing 昀椀nancial assistance to them can facilitate their re-entry into school (Baird et al., 2010; Barham et al., 2024). A recent study by Stenzel et al. (2024) focusing on schoolgirls in Ghana shows that the senior high school completion rate was higher among schoolgirls in districts where uptake of the FSHS policy was higher. Overall, for both boys and girls, completion rate was also higher in higher uptake districts. In this study, we set out to understand the implications of teenage birth on the future educational attainment of the mother. A common problem in empirically estimating this relationship is the problem of endogeneity of teen birth. We address this issue by instrumenting for teen birth using variables that capture both a physiological character- istic and the environmental conditions that can in昀氀uence teen births. Furthermore, we look into whether the FSHS policy moderates the impact of teen births on the educational attainment of young mothers. We then unpack the potential geographical differences in our objective by disaggregating our analysis for urban and rural localities where ed- ucation access and exposure to teenage pregnancy vary signi昀椀cantly. The remaining sections of the paper are organized as follows. Section two presents a review of the literature. In section three, we present the methods used in the analysis while the results and discussions are pre- sented in section four. Section 昀椀ve presents the summary and conclusion of the paper. 2. Literature review The focus of the current study is inspired by the human capital theory attributed to Becker (1964). In this theory, human capital is de昀椀ned as knowledge, skill and related attributes that affect human capabilities to do productive work (Becker, 1964; Schultz, 1984). Contrary to the view that economic differences between countries and among people is associated with variations in physical capital, Becker (1964) explains that the differences are in the variations in the level of human capital. Investment in human capital may take the form of medical care, schooling, on-the-job training, and obtaining information about the economic system such that the physical and mental capabilities of people are improved, thereby raising the prospect of real income. The theory emphasizes the relevance of education attainment to the wellbeing of an individual and community at large. This underscores the importance of the theory to the current study. Individuals with higher educational attainments are more likely to have better human capital that would improve economic outcomes. This implies that a teenage girl who drops out of school due to pregnancy or teen birth may have lower chances of improving their human capital and ultimately, their eco- nomic outcomes. For individuals from poor households, a girl’s inability to complete education because of early childbirth may have an impov- erishing effect, further perpetuating poverty cycle in the household. Over the years there has been a growing interest in the empirical literature to understand the determinants and implications of teenage birth. In the United States, Klepinger et al. (1999), using an instrumental variable technique, did not 昀椀nd suf昀椀cient evidence to accept the hy- pothesis that teenage pregnancy had a negative impact on educational attainments. Contrary evidence was provided by Schulkind and Sandler (2019) who showed that in the United States of America, teen girls whose high school education was interrupted by pregnancy have considerably lower educational attainment. However, the early child- birth does not have any effect on their chance of working, their future earnings or the total income of their families. Ardington et al. (2014) showed that in rural South Africa, women who had their 昀椀rst birth in their teen years, were less likely to complete high school compared to women who had their 昀椀rst childbirth after their teenage years. The authors also showed that teen birth was associated with higher risk of mortality. In Ghana, there have also been studies that used qualitative tech- niques to answer similar research questions. These studies largely used responses from interviews on the impact of teen birth on future eco- nomic outcomes. Gyan (2013) conducted a qualitative study on the causes of teenage childbearing and its effect on the educational attain- ment of the girl child at Chorkor, a district of Accra, Ghana. The study identi昀椀ed school dropout as a determinant and consequence of teenage birth. The study revealed that over 80% of interviewed girls dropped out of school due to pregnancy related issues. Biney and Nyarko (2017) assessed the relationship between adolescent births using the 2007 Ghana Maternal Health Survey. Using ANOVA and multivariate re- gressions, the authors found that adolescent births were associated with lower years of schooling. In a study conducted in the Volta region of Ghana, Morgan et al. (2022) also found that 79% of teenage mothers had dropped out of school. Two important issues emerge from the literature that justi昀椀es the need for the current study and hence its contribution to the literature. First, while the burden of teen birth is largest in sub-Saharan Africa, studies that investigate its impact and related pathways in the region are rather scant. Secondly, for the few studies that exist on Sub-Saharan Africa, the focus has been on qualitative techniques. Our study there- fore stands out in the way we attempt to correct for potential endoge- neity problems. This allows us to report robust estimates purged from endogeneity biases. We also disaggregate the analysis to identify various impact dynamics across geographical (rural/urban) locations in Ghana. We also assess how a national universal education policy (FSHS) could potentially mitigate the impact of teen birth on education outcomes through facilitating entry or re-entry into school for teen mothers. 3. Methodology Our empirical model is based on previous theoretical model pro- 2 https://ghana.unfpa.org/en/publications/昀椀ve-year-strategic-plan-addre ss-adolescent-pregnancy-ghana-2018-2022. 3 https://freeshs.gov.gh/. K.G. Tabiri et al. https://ghana.unfpa.org/en/publications/five-year-strategic-plan-address-adolescent-pregnancy-ghana-2018-2022 https://ghana.unfpa.org/en/publications/five-year-strategic-plan-address-adolescent-pregnancy-ghana-2018-2022 https://freeshs.gov.gh/ Social Sciences & Humanities Open 11 (2025) 101311 3 posed by Klepinger et al. (1999). The framework contrasts human cap- ital investment decisions for women who had their 昀椀rst child in their teen age and those who did not have children in their teen age. Against this backdrop, we examined the impact of teenage childbearing on the mother’s educational outcome by estimating the following equation yi =α + β1Teenbirth + β2Residence + β3Wealth status + γiXiH + εi (1) where yij is maternal educational outcome, Teenbirth is a dummy vari- able that equals one if the woman had a teenage birth and zero other- wise, Residence captures whether the woman lives in a rural or urban area, and Wealth status indicates the economic status of the woman’s household. X is a vector of other household and location characteristics (region of residence, ethnicity, sex of household head). To estimate how the FSHS policy mediates the impact of teen birth on the mother’s education, we estimate the following model yij =α + β1Teenbirth + β2FSHS + β3Teebirth × FSHS + β4Residence + β5Wealth status + γiXiH + εij (2) The primary parameter of interest in equation (2) is β3. It captures the impact of teen birth on the educational outcomes of mothers who were exposed to the FSHS. If this parameter is not different from zero, then the impact of teen birth on educational outcomes is the same for both bene昀椀ciaries and non-bene昀椀ciaries. However, if it is positive and β1 (the impact of teen birth) is negative, then the policy attenuates the impact of teen birth, and the overall impact of teen birth is given by β1 + β3. 3.1. Identi昀椀cation and estimation strategy A potential threat to the validity of the estimates from equation (1) is the problem of endogeneity. The decision to have a child during a girl’s teenage years may depend on several factors that could be peculiar to the girl. These factors may range from observables such as the socio- economic situation of the household or community to unobservable factors such as the girl’s ambitions or motivation to delay childbirth. Where the factors are observable, they can be included in the model to capture their effects. The challenge arises where these factors are not observable. This means that there could be potential correlation with the error term, and this can bias our estimates from the OLS regression. Another possible source of endogeneity is the problem of reverse cau- sality. This occurs where the dependent variable (education outcome) has a feedback effect on the key explanatory variable (teen birth). However, in this study, we do not envisage the latter source of endo- geneity. This is because of the time lag in the two variables. While teen birth may in昀氀uence future education outcomes, we cannot expect future education outcomes to in昀氀uence teen birth. Against this backdrop, we rule out this source of endogeneity in the current study. To address the challenge of endogeneity, we used two instrumental variables (IVs). Ideally, the IVs should be variables that are correlated with teen birth but only correlated with maternal education through teen birth. The 昀椀rst instrument we use is age at menarche, following Ribar (1994). We created a dummy variable that assigns a value of one to all women who had their 昀椀rst menstrual period before the median age at menarche, and zero otherwise. The intuition is that age at menarche is a predictor of the timing of fecundity and socio-sexual behavior (Presser, 1978; Ribar, 1994). Therefore, among sexually active girls, those with earlier age at menarche may be more likely to be pregnant and give birth in their teenage years. We also did not envisage any direct relationship with a woman’s future education outcomes except through teenage pregnancy and birth. Indeed, the age at menarche is a natural occur- rence and is therefore largely exogenous. The second instrument is the proportion of teen births in a cluster. This is the mean of the teen birth dummy variable for that cluster. A number of African studies such as Were (2007) and Mkhwanazi (2010) highlight the fact that social norms and pressures can predispose girls to teenage pregnancy. This instrumental variable therefore captures the predisposition to teenage pregnancy as a result of the girl’s social environment. While the IVs are intuitively appealing, we further test the strength and validity of the instrument using various statistical techniques. Speci昀椀cally, we use the Kleibergen-Paap rk Lagrange Multiplier test for underidenti昀椀cation, which accounts for heteroscedasticity, the Kleibergen-Paap rk F test for weak identi昀椀cation, and the Hansen J statistic to test for overidenti昀椀cation. The null hypothesis of the Sargan-Hansen test of overidentifying restrictions is that the instruments are valid, that is, they are uncorre- lated with the error term. Underidenti昀椀cation tests are conducted to determine whether the instruments provide suf昀椀cient information to identify the endogenous variable. If there is underidenti昀椀cation, the model cannot be estimated. The null hypothesis of the Kleibergen-Paap rk Lagrange Multiplier test is that the model is underidenti昀椀ed. We have weak identi昀椀cation when, even though the instruments are valid, they are only weakly correlated with the endogenous regressor. The Kleibergen-Paap rk F test is compared to critical values provided by Stock and Yogo (2005). A high test statistic (relative to the Stock-Yogo critical values) suggests that the instrument is weak. Taking the above into account, the IV estimation in this study is done using a two-stage process. The 昀椀rst stage estimates the following model: Teenbirthi =α + β1Z + γiXiH + εi (3) Where Z is made up of the two instruments (age at menarche and pro- portion of teen births) and X is a vector of other household and location characteristics. In the second stage, the predicted values of Teenbirth from (3) are used to estimate models (1) and (2). 3.2. Data and variables In this paper we use data from the most recent round of the Ghana Demographic and Health Survey (GDHS) conducted in 2022 by the Ghana Statistical Service (GSS). The GDHS, is a nationally representative survey conducted in Ghana within 昀椀ve (5) year intervals starting from 1988. The survey was designed to analyze, gather and publicize infor- mation on household characteristics, marriage, fertility, family plan- ning, reproductive health, child health, nutrition, mortality, contraception etc. In all, 618 clusters were selected for the 2022 GDHS. The sampling frame used was the updated frame prepared by the GSS based on the 2021 Population and Housing Census. In total, 18,450 households were selected for the survey, resulting in 15,014 interviewed women aged 15–49. For the purpose of this study, a total of 4091 women were sampled. This sample was obtained by 昀椀rst restricting the total sample to women who had ever given birth. In order to assess the impact of the FSHS policy, the sample was further restricted to women who were 30 years or younger at the time of the survey. In the Ghana Youth Survey by Du昀氀o et al. (2021), the minimum baseline age for students about to enter se- nior high school (SHS) was 13, with 80% being 18 and below, and 91% aged 19 and below. The 2012/2013 SHS Census report from the GSS also surveyed students between 15 and 18 (Ministry of Education, 2013). Based on these statistics, we estimated the average senior high school starting age to vary between 13 and 18 years. Women who 昀椀t this cri- terion (18 years old and below when the policy was introduced) were categorized as those exposed to the policy. These women would be 23 years old or below at the time of the 2022 GDHS. The women who were not exposed to policy are those who were already too old at the time of the introduction of the policy to have been bene昀椀ciaries. To ensure that these women were not too far removed from the policy, we limit the comparison group to those who are above 23 but at most 30 years old at the time of the 2022 GDHS. These women generally would have completed their senior high school studies prior to the implementation of the policy. Restricting it to this age range is K.G. Tabiri et al. Social Sciences & Humanities Open 11 (2025) 101311 4 therefore intended to help designate a more relevant comparison group for women who were potential bene昀椀ciaries of the FSHS policy (see Table 1). 4. Results 4.1. Summary statistics Summary statistics for key variables are presented in Table 2. The average number of years of education in the sample is about 8.6 years, with a maximum of 18 years of education, and some women having no education. Also, about 49.7% of the women in the sample reported having their 昀椀rst child in their teenage years. Thus, half of the women aged 30 or less who had ever given birth had their 昀椀rst child as teen- agers. Women dwelling in rural areas form a majority of the sample (57.5%). The proportion of women from poor households was 54.8% of the sample, with those from rich households accounting for about 24.7%. About 42% of the women in the sample had their 昀椀rst menses when they were less than 15 years old (15 is the median age at menarche in the sample). Women who were potential bene昀椀ciaries of the FSHS policy form 33% of the sample. 4.2. Teen birth and maternal education Table 3 shows the OLS and IV estimates for the effect of teen birth on maternal education. The OLS estimates are shown in columns (1) and (3), while the IV estimates are shown in columns (2) and (4). The stars on the Kleibergen-Paap rk Lagrange Multiplier test statistic for under- identi昀椀cation indicates that the p-value is less than 0.001, and therefore we reject the null hypothesis of underidenti昀椀cation. The stars on the Kleibergen-Paap rk F weak identi昀椀cation test statistic for under- identi昀椀cation indicates that it is higher than the Stock-Yogo critical values. The Hansen J statistic has no stars, indicating that it has a p- value greater than 0.10, and thus we do not reject the null hypothesis that the instruments are valid. These interpretations apply to subsequent results tables. The IV estimates for the coef昀椀cient of teen birth and that of the interaction term are both higher compared to the OLS estimates. This gives some indication that using the OLS would have underestimated the effect of teen birth on the maternal education outcome. We see from the results that women who had their 昀椀rst child in their teenage years have signi昀椀cantly lower years of education compared to their counterparts who had their 昀椀rst child later, all other things equal. Speci昀椀cally, we see from the IV regression in column (3) that women who had their 昀椀rst child as teenagers had about 2.4 less years of education compared to those who had their 昀椀rst child later. The positive and statistically sig- ni昀椀cant coef昀椀cient on the interaction between teen birth and being exposed to the FSHS policy also offers indication that the policy helps to attenuate the negative impact of teen birth on maternal education. On the average, women who had teen births but were exposed to the FSHS policy had 1.5 (3.175–1.707) years lower education compared to women who had no teen births. However, those who had teen births but were not exposed to the policy had 3.2 years less education, on average, compared to those who had no teen births. Other important factors associated with educational outcomes of the women in the sample include household wealth, sex of the household head, and the type of residence. Speci昀椀cally, we 昀椀nd that women from rich and middle-class households tend have more years of education compared to their counterparts from poor households, all else equal. Also, women from male-headed households had lower years of educa- tion compared to those from female-headed households, and women from rural households tend to have lower years of education compared to women living in urban areas, all other things equal. We found no statistically signi昀椀cant difference in years of education between Akan women and Ga/Dangme and Ewe women. However, our results showed Table 1 Variable description. Variable names Description Maternal Education Education in single years measured as completed years of schooling Teen birth Women who had their 昀椀rst birth at age 19 or below were assigned a value of 1 to mean “yes”, while women who had their 昀椀rst child after age 19 were given a value of 0 to mean “no” category. Age at menarche The age at which a woman had her 昀椀rst menstruation. Cluster teen birth average The proportion of women in a cluster who are assigned 1 for the teen birth variable FSHS exposure Women who were below 24 years at the time of the survey were considered potential bene昀椀ciaries and assigned a value of 1. Those aged 24–30 years were assigned a value of 0. Wealth The wealth index is an ordinal variable coded into 昀椀ve groups: poorest, poorer, middle, richer and richest. The study, however, re-classi昀椀es this variable into three groups representing poor, middle and rich respectively. Residence The variable captures whether the respondent was interviewed in a rural or urban location. 0 represents urban, while 1 represents rural. Region Woman’s administrative region of residence Ethnicity Woman’s ethnicity Sex of household head Equals 1 if male, 0 if female Source: Authors Table 2 Descriptive statistics. Variable N Mean Std. Dev. Min Max Education in single years 4091 8.559 5.143 0 18 Teen birth 4091 0.497 0.500 0 1 FSHS exposure 4091 0.330 0.470 0 1 Age 4091 25.1 3.607 15 30 Early menarche 4091 0.417 0.493 0 1 Cluster teen birth 4091 0.4974 0.2207 0 1 Wealth category Poor 4091 0.548 0.498 0 1 Middle 4091 0.205 0.404 0 1 Rich 4091 0.247 0.432 0 1 Ethnicity Akan 4091 0.307 0.461 0 1 Ga/Dangme 4091 0.035 0.185 0 1 Ewe 4091 0.088 0.284 0 1 Mole-Dagbani 4091 0.299 0.458 0 1 Gurma 4091 0.123 0.329 0 1 Others 4091 0.147 0.354 0 1 Residence Urban 4091 0.425 0.494 0 1 Rural 4091 0.575 0.494 0 1 Region Western 4091 0.052 0.223 0 1 Central 4091 0.059 0.236 0 1 Greater Accra 4091 0.044 0.205 0 1 Volta 4091 0.039 0.193 0 1 Eastern 4091 0.050 0.218 0 1 Ashanti 4091 0.066 0.249 0 1 Western North 4091 0.052 0.222 0 1 Ahafo 4091 0.055 0.228 0 1 Bono 4091 0.051 0.219 0 1 Bono East 4091 0.069 0.253 0 1 Oti 4091 0.068 0.252 0 1 Northern 4091 0.088 0.284 0 1 Savannah 4091 0.081 0.273 0 1 North East 4091 0.089 0.285 0 1 Upper East 4091 0.078 0.268 0 1 Upper West 4091 0.059 0.236 0 1 Source: Authors’ computation K.G. Tabiri et al. Social Sciences & Humanities Open 11 (2025) 101311 5 that women from the Mole-Dagbani, Gurma, and other northern ethnic groups tend to have lower years of education compared to Akans. Table 4 shows disaggregated results for teen mothers across types of residence (rural or urban). The disaggregation was necessary as girls face different social and economic conditions across these locations. The results largely re昀氀ect the evidence in Table 3 that teen births are asso- ciated with lower years of education, and being a potential FSHS ben- e昀椀ciary helps to reduce the negative impact. However, the coef昀椀cient of the interaction term is not statistically signi昀椀cant for the rural sample. In the context of this study, this implies that all other things equal, in rural areas, exposure to the FSHS policy makes no difference as far as the educational outcomes of teenage mothers are concerned. 5. Discussion and conclusion Teen births are a critical issue that has signi昀椀cant implications for the future of young girls, particularly in developing countries. This study set out to understand how teen births affect the educational attainment of the mother using the latest DHS data from Ghana. We further examined whether and how a policy that reduces the 昀椀nancial cost of senior high school education could moderate the effect of teen birth on educational attainment. Finally, we disaggregate our analysis based on location to study how the effect of the policy could depend on respondent’s place of residence. Our 昀椀ndings highlight the negative impact of teen births on educa- tional attainment for girls in Ghana. This corroborates the 昀椀ndings of several studies both in developing countries in Africa and in parts of the developed world (Kane et al., 2013; Kassahun et al., 2019; Sobngwi-- Tambekou et al., 2022). Furthermore, it highlights the fact that teenage pregnancy remains an important issue requiring renewed and sustained efforts from policymakers in order to solve. The 2022 GDHS report re- veals that 15% of women aged 15–19 years had ever been pregnant, and 11% had had a live birth (GSS and ICF, 2024). This shows the need for further strategic action to protect adolescent girls from early pregnancy and its potentially devastating effects. We also 昀椀nd that the FSHS policy attenuates the negative impact of teen birth on the educational attainment of mothers. This 昀椀nding has interesting implications. Generally, it highlights the importance of removing or reducing obstacles to the entry or re-entry of teenage mothers into school. However, it is critical to note that this 昀椀nding points to the fact that the FSHS policy helps to reduce the 昀椀nancial barrier to entry or re-entry into secondary education for teen mothers but does not eliminate it. Teen mothers, particularly those from less economically endowed homes, may still need to prioritize contributing to household income in order to care for their child (Morgan et al., 2023). A household’s economic burden increases with a teen mother and an additional child. Adding the cost of reintegrating the teen mother to school could be a dif昀椀cult 昀椀nancial decision. In this regard, our 昀椀ndings suggest that the FSHS policy could provide cushion for such households and encourages school reintegration for the mother. In addition to 昀椀nancial barriers, there are other factors such as social stigma, a lack of child-care facilities and support for the young mothers. Addressing these socioeconomic challenges are therefore crucial for ensuring that teen mothers are still able to pursue and achieve their educational goals and aspirations. The GES’s ‘Guidelines for Prevention of Pregnancy among School Girls and Facilitation of Re-entry into School after Childbirth’ provides a useful framework for addressing the educational challenges of teen mothers. The Guidelines emphasize creating awareness and providing social and psychological support to teen mothers as the primary ways of facilitating re-entry. Between the 2017/2018 and the 2019/2020 academic year, this policy helped 10,869 out of 22,147 pregnant girls (49.1%) in public schools to return to school (Africa Education Watch, 2022). This shows that while the guidelines can be helpful, there remains more to be done. The guidelines still face implementation challenges, including staf昀椀ng de昀椀cits, inade- quate funding, and the absence of a policy framework for providing material support for the young mothers (Africa Education Watch, 2022). Ensuring adequate and dedicated budgetary allocation will therefore be crucial going forward. Our evidence also con昀椀rms the need for policies like the re-entry program to be targeted for speci昀椀c populations. For instance, we found that for girls in rural areas, the FSHS policy was not immediately effective in mitigating the impact of teen birth on education. This may be due to a number of reasons. Educational attainment among Ghana’s rural populace is generally low, and especially so for females. According to the 2021 Population and Housing Census, about 35% of females aged 6 and older in rural areas had never attended school, compared to the national average of 24.5%. Furthermore, about 45% of females aged 6 and older in rural areas were not literate, as compared to the national average of 34.4% (GSS, 2023). Given this and the fact that the policy is relatively young, it may be dif昀椀cult to 昀椀nd signi昀椀cant impact in rural areas. Another reason could be the implementation challenges faced by the FSHS program. Gruijters et al. (2024) and Casely-Hayford et al. (2024) provide excellent review of free secondary policies in sub-Saharan Africa, highlighting various challenges in formulating as well as implementing such policies, including 昀椀scal sustainability and infrastructural challenges. Some of these issues are prevalent in Ghana’s case. There are concerns such as increased workload for existing teachers, inadequate classroom infrastructure, lack of teaching/learning materials, and delay in the supply of funds contribute to exacerbating Table 3 Effect of teen birth on education of teen mothers. OLS IV (1) (2) (3) (4) Teen birth −1.516*** −2.166*** −2.361*** −3.175*** (0.136) (0.171) (0.316) (0.403) FSHS ​ 0.064 ​ 0.125 ​ (0.228) ​ (0.442) Teen birth * FSHS ​ 1.391*** ​ 1.707** ​ (0.293) ​ (0.669) Wealth Middle 1.494*** 1.573*** 1.440*** 1.539*** (0.194) (0.193) (0.197) (0.195) Rich 3.167*** 3.284*** 2.985*** 3.134*** (0.204) (0.204) (0.212) (0.210) Male household head −0.626*** −0.598*** −0.616*** −0.580*** (0.146) (0.145) (0.146) (0.146) Ethnicity Ga/Dangme −0.220 −0.234 −0.265 −0.285 (0.342) (0.343) (0.346) (0.348) Ewe −0.381 −0.352 −0.413 −0.378 (0.265) (0.262) (0.266) (0.264) Mole-Dagbani −1.574*** −1.529*** −1.594*** −1.538*** (0.240) (0.239) (0.239) (0.239) Gurma −2.524*** −2.519*** −2.572*** −2.564*** (0.311) (0.308) (0.313) (0.309) Others −1.336*** −1.267*** −1.353*** −1.265*** (0.257) (0.257) (0.257) (0.258) Rural −0.757*** −0.727*** −0.728*** −0.690*** (0.161) (0.160) (0.161) (0.160) Constant 10.257*** 10.083*** 10.806*** 10.582*** (0.303) (0.306) (0.359) (0.353) N 4091 4091 4091 4091 R-Squared 0.302 0.312 0.296 0.305 Adjusted R-Squared 0.298 0.308 0.292 0.300 Underidenti昀椀cation ​ ​ 696.843*** 388.341*** Weak identi昀椀cation ​ ​ 649.580*** 164.909*** Overidenti昀椀cation ​ ​ 0.001 0.794 *p < 0.10, **p < 0.05, ***p < 0.01. All regressions include region dummies. Robust standard errors in parentheses. K.G. Tabiri et al. Social Sciences & Humanities Open 11 (2025) 101311 6 the already dif昀椀cult circumstance that senior high schools in rural areas in Ghana face (Chanimbe & Dankwah, 2021). This calls for carefully tailored interventions for these groups of people with peculiar chal- lenges including relatively poor infrastructure, economic conditions and cultural practices/beliefs that facilitate risky sexual behavior. While our study addresses an important social concern in developing countries including Ghana, it also has some limitations worth mentioning. First, the FSHS policy is only a few years old. Therefore, the full scale of impact or its mitigating effect may be yet to be fully realized. There is also the empirical issue of clearly identifying a counterfactual given the absence of a randomized control trial (RCT), widely consid- ered to be ‘gold standard’. Since the policy was universally implemented by the government, such data is not available. Stenzel (2024) used up- take intensity from the Educational Management Information System by the Ministry of Education. However, this data is no longer publicly available. While we do our best to 昀椀nd a suf昀椀ciently appropriate com- parison group, we acknowledge that RCT data would have provided a more precise estimate of the impact of teen birth and of the FSHS policy. It is also important to recognize that the FSHS variable we use measures exposure and not actual uptake of the policy. Our 昀椀ndings are, however, relevant in the several ways they contribute to the already scant liter- ature on the subject matter as earlier indicated. Finally, we acknowledge that there may be disparities within rural and urban settings, the exploration of which would provide even further insight into the effect of teen birth on educational outcomes and the mediating role of edu- cation. The data we use for this study is however limited in this regard. Exploring these within-locality differences would be a useful starting point for future studies in this area. 6. Conclusion Teen births pose signi昀椀cant challenges to the educational attainment of girls and subsequently affects other aspects such as labor market outcomes and economic empowerment. The results of this study provide empirical evidence of how reducing the 昀椀nancial barriers can lessen the negative effect of teen birth on the educational attainment of girls in Ghana. It also highlights how the peculiar circumstances of rural areas in Ghana can immune them from the bene昀椀ts of otherwise useful policies. This study therefore emphasizes the need to bolster efforts to further reduce all kinds of barriers to the re-entry of teen mothers into school by facilitating the proper implementation of the existing guidelines, as well as implementing additional measures that are tailored to the peculiar needs of rural areas. CRediT authorship contribution statement Kwasi Gyabaa Tabiri: Writing – review & editing, Writing – original draft, Software, Methodology, Formal analysis, Conceptualization. Muniratu Tahiru: Writing – original draft, Methodology, Conceptuali- zation. Yaw Boateng Atakorah: Writing – original draft, Methodology. Jacob Novignon: Writing – review & editing, Writing – original draft. Table 4 Effect of teen birth on maternal education (by locality). OLS IV (1) (2) (3) (4) (5) (6) (7) (8) Rural Urban Rural Urban Rural Urban Rural Urban Teen birth −1.040*** −2.181*** −1.774*** −2.628*** −2.105*** −2.668*** −2.645*** −3.442*** (0.187) (0.194) (0.233) (0.250) (0.479) (0.433) (0.612) (0.553) FSHS ​ ​ 0.491 −0.227 ​ ​ 1.212* −0.476 ​ ​ (0.315) (0.312) ​ ​ (0.621) (0.604) Teen birth * FSHS ​ ​ 1.307*** 1.154*** ​ ​ 0.607 1.895** ​ ​ (0.395) (0.423) ​ ​ (0.930) (0.947) Wealth Middle 1.398*** 1.391*** 1.443*** 1.466*** 1.321*** 1.368*** 1.365*** 1.477*** (0.263) (0.302) (0.260) (0.303) (0.268) (0.303) (0.264) (0.302) Rich 2.776*** 3.311*** 2.912*** 3.394*** 2.525*** 3.216*** 2.713*** 3.321*** (0.316) (0.290) (0.318) (0.291) (0.334) (0.299) (0.333) (0.295) Male household head −0.697*** −0.639*** −0.643*** −0.625*** −0.695*** −0.631*** −0.627*** −0.608*** (0.204) (0.203) (0.203) (0.202) (0.204) (0.202) (0.202) (0.202) Ethnicity Ga/Dangme −0.438 0.065 −0.576 0.114 −0.627 0.099 −0.781 0.184 (0.529) (0.424) (0.527) (0.428) (0.542) (0.424) (0.543) (0.432) Ewe −0.502 −0.412 −0.533 −0.355 −0.554 −0.446 −0.593* −0.365 (0.350) (0.393) (0.345) (0.392) (0.352) (0.394) (0.347) (0.392) Mole-Dagbani −1.913*** −1.796*** −1.873*** −1.770*** −1.984*** −1.782*** −1.943*** −1.737*** (0.323) (0.375) (0.321) (0.375) (0.321) (0.372) (0.320) (0.373) Gurma −2.531*** −2.188*** −2.532*** −2.203*** −2.636*** −2.222*** −2.628*** −2.254*** (0.407) (0.495) (0.402) (0.493) (0.410) (0.494) (0.405) (0.492) Others −1.629*** −1.398*** −1.558*** −1.362*** −1.698*** −1.398*** −1.609*** −1.345*** (0.368) (0.357) (0.367) (0.358) (0.369) (0.356) (0.367) (0.359) Constant 10.242*** 9.882*** 9.990*** 9.800*** 10.924*** 10.218*** 10.476*** 10.162*** (0.404) (0.382) (0.404) (0.389) (0.491) (0.473) (0.485) (0.461) N 2353 1738 2353 1738 2353 1738 2353 1738 R-Squared 0.284 0.259 0.300 0.263 0.274 0.256 0.289 0.258 Adjusted R-Squared 0.276 0.248 0.292 0.252 0.266 0.246 0.281 0.247 Under identi昀椀cation ​ ​ ​ ​ 341.366*** 328.291*** 310.705*** 190.978*** Weak identi昀椀cation ​ ​ ​ ​ 280.751*** 344.286*** 125.081*** 88.394*** Over identi昀椀cation ​ ​ ​ ​ 0.066 0.522 0.821 0.907 *p < 0.10, **p < 0.05, ***p < 0.01. All regressions include region dummies. Robust standard errors in parentheses. K.G. Tabiri et al. Social Sciences & Humanities Open 11 (2025) 101311 7 Data availability The data used in the study was obtained from the Demographic and Health Surveys. The data are publicly available at https://www.dhspr ogram.com. Funding declaration This research did not receive any speci昀椀c grant from funding agencies in the public, commercial, or not-for-pro昀椀t sectors. Declaration of competing interest The authors declare that they have no known competing 昀椀nancial interests or personal relationships that could have appeared to in昀氀uence the work reported in this paper. Acknowledgments We are grateful to the Ghana Statistical Service and the DHS program for providing access to the data used in the study. References Africa Education Watch. (2022). Re-entry of pregnant girls and teenage mothers to school. https://healtheducationresources.unesco.org/library/documents/re-entr y-pregnant-girls-and-teenage-mothers-school-critical-policy-and-strategy. Ardington, C., Menendez, A., & Mutevedzi, T. (2014). Early childbearing, human capital attainment, and mortality risk: Evidence from a longitudinal demographic surveillance area in rural kwazulu-natal, South Africa. Economic Development and Cultural Change, 63(2), 281–317. https://doi.org/10.1086/678983 Baird, S., Chirwa, E., McIntosh, C., & Özler, B. (2010). The short-term impacts of a schooling conditional cash transfer program on the sexual behavior of young women. Health Economics, 19, 55–68. https://doi.org/10.1002/hec.1569 Barham, T., Macours, K., & Maluccio, J. A. (2024). Experimental evidence from a conditional cash transfer program: Schooling, learning, fertility, and labor market outcomes after 10 years. Journal of the European Economic Association. https://doi. org/10.1093/jeea/jvae005 Becker, G. S. (1964). Human capital: A theoretical and empirical analysis, with special reference to education (3rd ed. ed.). Chicago: University of Chicago press. Biney, A. A. E., & Nyarko, P. (2017). Is a woman’s 昀椀rst pregnancy outcome related to her years of schooling? An assessment of women’s adolescent pregnancy outcomes and subsequent educational attainment in Ghana. Reproductive Health, 14, 123. https:// doi.org/10.1186/s12978-017-0378-2 Casely-Hayford, L., Gruijters, R. J., Adjei, L. O., & Yeboah, V. A. (2024). The intended and unintended effects of secondary school fee abolition: Evidence from Ghana’s free senior high school policy. SocArXiv. https://doi.org/10.31235/osf.io/hsp7b. April 10. Chandra-Mouli, V., Mccarraher, D. R., Phillips, S. J., Williamson, N. E., & Hainsworth, G. (2014). Contraception for adolescents in low and middle income countries: Needs, barriers, and access. Reproductive Health, 11(1), 1–8. https://doi.org/10.1186/1742- 4755-11-1 Chanimbe, T., & Dankwah, K. O. (2021). The ‘new’ free senior high school policy in Ghana: Emergent issues and challenges of implementation in schools. Interchange, 52, 599–630. https://doi.org/10.1007/s10780-021-09440-6 Du昀氀o, E., Dupas, P., & Kremer, M. (2021). The impact of free secondary education: Experimental evidence from Ghana. http://www.nber.org/papers/w28937. Ghana Education Service. (2018). Guidelines for prevention of pregnancy among school girls and facilitation of Re-entry into school after childbirth. https://features.hrw.or g/features/african-union/files/Ghana%20-%202018%20Guidelines%20for%20Pre vention%20of%20Pregnancy%20Among%20School%20Girls%20and%20Facilitat ion%20of%20Re-Entry%20into%20School%20After%20Childbirth.pdf. Ghana Statistical Service. (2023). International day of rural women press release. https ://www.statsghana.gov.gh/gssmain/昀椀leUpload/pressrelease/International%20Day %20of%20Rural%20Women%20Press%20Release%20from%20GSS.pdf. Groves, A. K., Gebrekristos, L. T., McNaughton Reyes, L., Moodley, D., Raziano, V., & Maman, S. (2022). A mixed-methods study of resilience and return to school among adolescent mothers in South Africa. Global Public Health, 17(9), 2111–2124. https:// doi.org/10.1080/17441692.2021.1970208 Gruijters, R. J., Abango, M. A., & Casely-Hayford, L. (2024). Secondary school fee abolition in sub-Saharan Africa: Taking stock of the evidence. Comparative Education Review, 68(3), 370–402. https://doi.org/10.1086/731496 Gyan, C. (2013). The effects of teenage pregnancy on the educational attainment of girls at chorkor, a suburb of Accra. Journal of Educational and Social Research, 3(3), 53–60. https://doi.org/10.5901/jesr.2013.v4n3p53 Gyesaw, N. Y. K., & Ankomah, A. (2013). Experiences of pregnancy and motherhood among teenage mothers in a suburb of Accra, Ghana: A qualitative study. International Journal of Women’s Health, 2013(5), 773–780. https://doi.org/ 10.2147/IJWH.S51528 Kane, J. B., Morgan, S. P., Harris, K. M., & Guilkey, D. K. (2013). The educational consequences of teen childbearing. Demography, 50(6), 2129–2150. https://doi.org/ 10.1007/s13524-013-0238-9 Kassahun, E. A., Gelagay, A. A., Muche, A. A., et al. (2019). Factors associated with early sexual initiation among preparatory and high school youths in woldia town, northeast Ethiopia: A cross-sectional study. BMC Public Health, 19, 378. https://doi. org/10.1186/s12889-019-6682-8 Klepinger, D., Lundberg, S., & Plotnick, R. (1999). How does adolescent fertility affect the human capital and wages of young women? Journal of Human Resources, 34(3), 421–423. https://doi.org/10.2307/146375 Kyei-Arthur, F., Agyekum, M. W., & Kyei-Gyam昀椀, S. (2024). “You cannot stay with one person once you begin having sex at a young age”: The prevalence, correlates and effects of early sexual debut among children in Ghana. Reproductive Health, 21(38). https://doi.org/10.1186/s12978-024-01775-4 Ministry of Education. (2013). Report on basic statistics and planning parameters for senior high schools in Ghana - 2012/2013. https://microdata.statsghana.gov.gh /index.php/catalog/69/download/266. Mkhwanazi, N. (2010). Understanding teenage pregnancy in a post-apartheid South African township. Culture, Health and Sexuality, 12(4), 347–358. https://doi.org/ 10.1080/13691050903491779 Morgan, A. K., Agyemang, S., Dogbey, E., Arimiyaw, A. W., & Owusu, A. F. S. (2022). “We were girls but suddenly became mothers”: Evaluating the effects of teenage motherhood on girl’s educational attainment in the Volta Region. Cogent Social Sciences, 8(1), Article 2036312. https://doi.org/10.1080/23311886.2022.2036312 Morgan, A. K., Ibrahim, R., Owusu, A. F. S., Awafo, B. A., Quartey, T., Aziire, M. A., … Katey, D. (2023). Prevalence and associated factors of school re-entry among teenage mothers in Ghana’s Volta region: A cross-sectional survey. International Journal of Adolescence and Youth, 28(1). https://doi.org/10.1080/02673843.2023.2242476 Neal, S., Channon, A., Chandra-Mouli, V., et al. (2020). Trends in adolescent 昀椀rst births in sub-saharan Africa: A tale of increasing inequity? International Journal for Equity in Health, 19(151). https://doi.org/10.1186/s12939-020-01251-y Presser, H. B. (1978). Age at menarche, socio-sexual behavior, and fertility. Social Biology, 25(2), 94–101. https://doi.org/10.1080/19485565.1978.9988327 Ribar, D. C. (1994). Teenage fertility and high school completion. The Review of Economics and Statistics, 76(3), 413–424. https://doi.org/10.2307/2109967 Schulkind, L., & Sandler, D. H. (2019). The timing of teenage births: Estimating the effect on high school graduation and later-life outcomes. Demography, 56(1), 345–365. https://doi.org/10.1007/s13524-018-0748-6 Schultz, T. P. (1984). Studying the impact of household economic and community variables on child mortality. Population and Development Review, 10, 215–235. https://doi.org/10.2307/2807962 Sobngwi-Tambekou, J. L., Tsague-Agnoux, M., Fezeu, L. K., et al. (2022). Teenage childbearing and school dropout in a sample of 18,791 single mothers in Cameroon. Reproductive Health, 19(10). https://doi.org/10.1186/s12978-021-01323-4 Stenzel, A. G., Osei Kwadwo, V., & Vincent, R. C. (2024). Free secondary education policy and education attainment. International Journal of Educational Development, 106, Article 103021. https://doi.org/10.1016/j.ijedudev.2024.103021 Stock, J. H., & Yogo, M. (2005). Testing for weak instruments in linear IV regression. In D. W. K. Andrews, & J. H. Stock (Eds.), Identi昀椀cation and inference for econometric models: Essays in honor of thomas rothenberg (pp. 80–108). Cambridge: Cambridge University Press, 2005. Working paper version: NBER Technical Working Paper 284 http://www.nber.org/papers/T0284. Were, M. (2007). Determinants of teenage pregnancies: The case of Busia District in Kenya. Economics and Human Biology, 5(2), 322–339. https://doi.org/10.1016/j. ehb.2007.03.005 WHO. (2023). Adolescent pregnancy. Retrieved from https://www.who.int/news-room/ fact-sheets/detail/adolescent-pregnancy. (Accessed 22 May 2024). WHO and UNICEF. (2023). Trends in maternal mortality: 2000-2020: Estimates by WHO, UNICEF, UNFPA, world bank group and the UNDESA/population division. Geneva: World Health Organization. https://www.who.int/publications/i/item/9789 240068759. World Bank. (2024). World development indicators. https://databank.worldbank.org/ source/world-development-indicators. (Accessed 22 May 2024). Yakubu, I., & Salisu, W. J. (2018). Determinants of adolescent pregnancy in sub-saharan Africa: A systematic review. Reproductive Health, 15(15). https://doi.org/10.1186/ s12978-018-0460-4 K.G. Tabiri et al. https://www.dhsprogram.com https://www.dhsprogram.com https://healtheducationresources.unesco.org/library/documents/re-entry-pregnant-girls-and-teenage-mothers-school-critical-policy-and-strategy https://healtheducationresources.unesco.org/library/documents/re-entry-pregnant-girls-and-teenage-mothers-school-critical-policy-and-strategy https://doi.org/10.1086/678983 https://doi.org/10.1002/hec.1569 https://doi.org/10.1093/jeea/jvae005 https://doi.org/10.1093/jeea/jvae005 http://refhub.elsevier.com/S2590-2911(25)00038-5/sref6 http://refhub.elsevier.com/S2590-2911(25)00038-5/sref6 https://doi.org/10.1186/s12978-017-0378-2 https://doi.org/10.1186/s12978-017-0378-2 https://doi.org/10.31235/osf.io/hsp7b https://doi.org/10.1186/1742-4755-11-1 https://doi.org/10.1186/1742-4755-11-1 https://doi.org/10.1007/s10780-021-09440-6 http://www.nber.org/papers/w28937 https://features.hrw.org/features/african-union/files/Ghana%20-%202018%20Guidelines%20for%20Prevention%20of%20Pregnancy%20Among%20School%20Girls%20and%20Facilitation%20of%20Re-Entry%20into%20School%20After%20Childbirth.pdf https://features.hrw.org/features/african-union/files/Ghana%20-%202018%20Guidelines%20for%20Prevention%20of%20Pregnancy%20Among%20School%20Girls%20and%20Facilitation%20of%20Re-Entry%20into%20School%20After%20Childbirth.pdf https://features.hrw.org/features/african-union/files/Ghana%20-%202018%20Guidelines%20for%20Prevention%20of%20Pregnancy%20Among%20School%20Girls%20and%20Facilitation%20of%20Re-Entry%20into%20School%20After%20Childbirth.pdf https://features.hrw.org/features/african-union/files/Ghana%20-%202018%20Guidelines%20for%20Prevention%20of%20Pregnancy%20Among%20School%20Girls%20and%20Facilitation%20of%20Re-Entry%20into%20School%20After%20Childbirth.pdf https://www.statsghana.gov.gh/gssmain/fileUpload/pressrelease/International%20Day%20of%20Rural%20Women%20Press%20Release%20from%20GSS.pdf https://www.statsghana.gov.gh/gssmain/fileUpload/pressrelease/International%20Day%20of%20Rural%20Women%20Press%20Release%20from%20GSS.pdf https://www.statsghana.gov.gh/gssmain/fileUpload/pressrelease/International%20Day%20of%20Rural%20Women%20Press%20Release%20from%20GSS.pdf https://doi.org/10.1080/17441692.2021.1970208 https://doi.org/10.1080/17441692.2021.1970208 https://doi.org/10.1086/731496 https://doi.org/10.5901/jesr.2013.v4n3p53 https://doi.org/10.2147/IJWH.S51528 https://doi.org/10.2147/IJWH.S51528 https://doi.org/10.1007/s13524-013-0238-9 https://doi.org/10.1007/s13524-013-0238-9 https://doi.org/10.1186/s12889-019-6682-8 https://doi.org/10.1186/s12889-019-6682-8 https://doi.org/10.2307/146375 https://doi.org/10.1186/s12978-024-01775-4 https://microdata.statsghana.gov.gh/index.php/catalog/69/download/266 https://microdata.statsghana.gov.gh/index.php/catalog/69/download/266 https://doi.org/10.1080/13691050903491779 https://doi.org/10.1080/13691050903491779 https://doi.org/10.1080/23311886.2022.2036312 https://doi.org/10.1080/02673843.2023.2242476 https://doi.org/10.1186/s12939-020-01251-y https://doi.org/10.1080/19485565.1978.9988327 https://doi.org/10.2307/2109967 https://doi.org/10.1007/s13524-018-0748-6 https://doi.org/10.2307/2807962 https://doi.org/10.1186/s12978-021-01323-4 https://doi.org/10.1016/j.ijedudev.2024.103021 http://www.nber.org/papers/T0284 https://doi.org/10.1016/j.ehb.2007.03.005 https://doi.org/10.1016/j.ehb.2007.03.005 https://www.who.int/news-room/fact-sheets/detail/adolescent-pregnancy https://www.who.int/news-room/fact-sheets/detail/adolescent-pregnancy https://www.who.int/publications/i/item/9789240068759 https://www.who.int/publications/i/item/9789240068759 https://databank.worldbank.org/source/world-development-indicators https://databank.worldbank.org/source/world-development-indicators https://doi.org/10.1186/s12978-018-0460-4 https://doi.org/10.1186/s12978-018-0460-4 Does free secondary education mitigate the impact of teen births on educational outcomes? Evidence from Ghana 1 Introduction 2 Literature review 3 Methodology 3.1 Identification and estimation strategy 3.2 Data and variables 4 Results 4.1 Summary statistics 4.2 Teen birth and maternal education 5 Discussion and conclusion 6 Conclusion CRediT authorship contribution statement Data availability Funding declaration Declaration of competing interest Acknowledgments References