Measurement and decomposition of education-related inequality in exclusive breastfeeding practice among Ethiopian mothers: applying Wagstaff decomposition analysis Authors: Anteneh Mengist Dessie, Melkamu Aderajew Zemene, Asaye Alamneh Gebeyehu, Denekew Tenaw Anley, Rahel Mulatie Anteneh, Natnael Moges, Ermias Sisay Chanie, Sintayehu Simie Tsega, Melaku Ashagrie Belete, Ermiyas Alemayehu, Natnael Kebede Published: 2024 ABSTRACT BackgroundHuman breast milk, a naturally balanced source of infant nutrition, promotes optimal growth and health when exclusively fed for 6 months. Exclusive breastfeeding reduces common childhood infections, provides protection against some chronic illnesses, and contributes to achieving several Sustainable Development Goals. Despite its benefits, only 58% of Ethiopian women practice it, and the associated education-related inequality is not well documented. Thus, this study aims to quantify and decompose the education-related inequality in exclusive breastfeeding practice among Ethiopian mothers.MethodsA total of 1,504 weighted samples were studied using a Performance Monitoring for Action Ethiopia longitudinal panel survey dataset (2021–2023). Wagstaff normalized concentration index and its concentration curve were used to assess education-related inequality in exclusive breastfeeding practice. Wagstaff decomposition analysis was performed to decompose the concentration index and identify factors contributing to the observed education-related inequality. Significance was declared at p-value <0.05.ResultsThe overall prevalence of exclusive breastfeeding among Ethiopian women was 57.29% (95% CI: 54.79, 59.80%), with a greater concentration found among women with lower levels of education. This indicates an inequality that favors less educated women (pro-less educated inequality), as demonstrated by the Wagstaff normalized concentration index of −0.058 (95% CI: −0.113, −0.002). Factors that made a significant contribution to the observed education-related inequality in exclusive breastfeeding practice were residence (18.80%), region (29.31%), place of birth (−7.38%), and the wantedness status of the indexed pregnancy (82.58%). The indexed pregnancy’s wantedness was made a more elastic (elasticity = 0.282) contribution.ConclusionThe study identified a small yet significant education-related inequality in exclusive breastfeeding, favoring less educated women. Hence, emphasis should be placed not only on educating women but also on healthy habits that they can leave behind when they learn. Residence, region, place of birth, and indexed pregnancy’s wantedness significantly contributed to the observed education-related inequality. The elasticity value for each factor suggests that policy changes addressing these factors could readily reduce the observed inequality. Introduction Human breast milk is naturally balanced nutritious food for infants. Really, it is an ideal food for a child’s survival, growth, and development. Exclusive breastfeeding (EBF) is the practice of providing infants with only breast milk, including expressed breast milk, for the initial 6 months of life, with no introduction of additional liquids or solid foods, except for necessary medications (1). The World Health Organization (WHO) and United Nations Children’s Fund (UNICEF) recommend that all mothers should breastfeed their children exclusively for the first 6 months (2). For almost all infants, it is the simplest, most efficient, healthiest and cost-effective intervention to fulfill their requirements. However, only about 44% of infants aged 0–6 months worldwide were exclusively breastfed over the period of 2015–2020 (3). Exclusive breastfeeding decreases the occurrence of common childhood illnesses such as diarrhea and acute respiratory infections. Moreover, it provides protection against some chronic illnesses and diseases, lowering the risk of asthma, obesity, type 1 diabetes, and sudden infant death syndrome (SIDS) in exclusively breastfed infants (4). Overall, it is estimated that EBF, combined with almost universal breastfeeding up to 1 year of age, has the potential to avert 13% of all deaths in children under the age of five in developing regions (5, 6). Beyond the direct health benefits for the child, exclusive breastfeeding plays a crucial role in achieving many of the Sustainable Development Goals (SDGs). These include SDG 2, addressing hunger and enhancing nutrition; SDG 3, reducing child mortality and mitigating the risk of non-communicable diseases; and SDG 4, promoting cognitive development and supporting education (7). The Ethiopian government has implemented various interventions aimed at enhancing the practice of exclusive breastfeeding, including health information dissemination through various channels, the development of training materials and implementation guidelines for breastfeeding (8, 9), and the integration of breastfeeding into the primary healthcare system in alignment with the health extension program (10). Despite the recognized benefits and numerous efforts to promote exclusive breastfeeding, the practice remains below the recommended level. While breastfeeding is common in Ethiopia, with 97% of infants receiving breast milk at some point, only 58% of mothers exclusively breastfeed their infants (11). Additionally, the progress in exclusive breastfeeding has been minimal, with rates increasing only from 54.5% (95% CI: 49.9–59.0%) in 2000 to 59.9% (95% CI: 55.0–64.5%) in 2016 (12). Various socio-demographic, economic, and obstetric factors influence the exclusive breastfeeding practices of women in Ethiopia (11, 13–15). Education is widely acknowledged as a key determinant of health outcomes, and its role in shaping exclusive breastfeeding practices has been explored in various contexts. Although highly educated women may have greater access to information about the benefits of exclusive breastfeeding, they often face challenges in practicing it due to the demands of their careers, which frequently require them to return to work soon after giving birth (13, 16). Additionally, concerns about body image among educated women (17) may further hinder their commitment to exclusive breastfeeding (18, 19). Therefore, this study aims to contribute to the existing literature by employing the Wagstaff decomposition analysis to measure and decompose education-related inequality in exclusive breastfeeding practices among Ethiopian women. This method allows for quantifying the extent of education-related inequality in exclusive breastfeeding practices, identifying the factors contributing to the observed inequality, and provides corresponding elasticities to elucidate how changes in the determinant relate to changes in the EBF practice. Methods and materials Study design, setting and period A longitudinal panel survey design was employed to generate data regarding education-related inequality in exclusive breastfeeding practice among Ethiopian mothers. Ethiopia, where this study was conducted, is situated in the Horn of Africa (30–150N latitude and 330–480E longitude). The country occupies an area of 1.1 million square kilometers with an altitude that ranges from the highest peak at Ras Dashen (4,620 m above sea level) down to the Dallol depression, about 148 m below sea level. It has nine regional states (Afar, Amhara, Benishangul Gumuz, Gambela, Hareri, Oromia, Somali, Southern Nations Nationalities and People (SNNP), and Tigray) and two city administrations (Addis Ababa and Dire Dawa). Data source, study populations, and sampling This study utilized the performance monitoring for action Ethiopia 2021 (PMA-ET 2021) longitudinal panel survey datasets. The PMA Ethiopia collected cross-sectional data annually from 2019 to 2023 and longitudinal data (following pregnant women through 1 year postpartum) in two cohorts of women: cohort 1 (2019–2021) and cohort 2 (2021–2023). The second longitudinal panel survey that we utilized in this specific study was conducted between 2021 and 2023 in Oromia, Amhara, SNNP regional states, and one urban region, Addis Ababa, that collectively represent 85% of the population in Ethiopia. The 2021 PMA Ethiopia panel survey employed two-stage stratified cluster sampling, where 162 enumeration areas (EAs) proportionally from both urban and rural strata and households (35 households per EA) were selected from the four regions at the first and second stages, respectively. Then, 2,298 pregnant women or less than 6 weeks postpartum in a given household were enrolled in the panel survey, and they were surveyed again after 6 weeks, 6 months, and a year postpartum. However, in this study, only 1,499 women (1,504 weighted samples) who participated in all three consecutive surveys (baseline, 6 weeks, and 6 months postpartum) and completed the interview were included in the final analysis. The 799 women who were dropped had a mean age of 26.93 years (SD = 6.65). Two hundred eighty-two (35.29%) of these women had identified as Orthodox, 280 (35.04%) as Protestant, 225 (28.16%) as Muslim, and 12 (1.52%) as belonging to other religions. They were from Amhara (144, 18.02%), Oromia (275, 34.42%), SNNP (271, 33.92%), and Addis Ababa (109, 13.64%). The majority, 91.61% (732), were married, with 41.93% (335) living in urban areas and 58.07% (464) in rural areas. Variables of the study The outcome variable for the study was exclusive breastfeeding which was coded as 0 for ‘not exclusively breastfeed’ and 1 for ‘exclusively breastfeed’. Women’s educational status was the measure of inequality in this study. Several sets of explanatory variables were used to explain the education-related inequalities in exclusive breastfeeding practice of women. Socio-demographic factors such as residence, wealth status, age, religion, marital status, and region, and pregnancy-related factors such as the wantedness status of the indexed pregnancy and place of delivery were all taken into account. The selection of these covariates was based on a review of the existing literature and their availability in our data source. Statistical analysis Measuring education-related inequalities in EBF practice All statistical analyses were conducted using Stata version 17. Two parameters were used to measure education-related inequality in EBF practice of women: (i) concentration curve (CC) and (ii) concentration index (C), using the highest level of school the women attended. The concentration curve plots the cumulative percentage of EBF practice (y-axis) against the cumulative percentage of women ranked by their educational status starting from the never attended school to the higher education (x-axis). There is an equality line (a 45°-line that extends from the lower left corner to the upper right corner) on the graph. A concentration curve above the line of equality indicates that EBF is disproportionately concentrated among women with low levels of education, and the reverse is true (20). The concentration index, a common measure of inequality in health, was calculated to quantify the extent of education-related inequalities in the EBF practice of women (21). The C is twice the area between the concentration curve and the line of equality (21, 22). It can be written as follows (23): Where hi is the health outcome (EBF) for the ith woman; μ is the mean of h; Ri is the ith-ranked individual in the educational status distribution from the most disadvantaged (i.e., never attended school) to the least disadvantaged (i.e., higher education); and n is the number of women. The concentration index value ranges from −1 to 1, where a negative value indicates that the EBF is disproportionately concentrated among women with low levels of education, and a positive value indicates that the EBF is disproportionately concentrated among women with high levels of education. If there is no inequality, it equals 0. However, when the outcome variable is a bounded variable (like EBF with values 0 and 1 in this study), the C will estimate the extent of inequality incorrectly. Thus, in this study, Wagstaff normalization was applied, and the Wagstaff normalized concentration index was calculated by dividing C by 1 minus the mean (1 – μ) (24). The statistical significance of the Wagstaff normalized concentration index was declared at a p-value <0.05. Decomposition of the concentration index Wagstaff decomposition analysis was performed to decompose the concentration index and explain the contribution of each determinants on the observed overall education-related inequality in the EBF practice of women (22). This decomposition method was first introduced with a linear model. However, an appropriate statistical technique (probit model) that yields probabilities in the range 0–1 was applied to adjust it for non-linear health variables (25). Thus, the decomposition of the C in this study is based on regression analysis (maximum-likelihood probit model) of the relationship between an outcome variable and a set of determinants. The overall concentration index comprises the explained part (the sum of the absolute contribution of all factors) and the unexplained part (the residual component). The final output of the Wagstaff decomposition process displays each factor’s coefficient, concentration index, its absolute and percentage contribution to the total observed education-related inequality, and the elasticity (the sensitivity of the outcome variable to that factor). Here, the regression coefficients reveal the relationship between the factor and the outcome variable (EBF), while the concentration index illustrates how the factor itself is distributed across the ranking variable (education). Elasticity is determined by weighting regression coefficients, giving higher weight to more prevalent determinants, and multiplying the result by the ratio of the mean of the determinant to the mean of the outcome. This elasticity provides insight into how changes in the determinant relate to changes in the outcome variable. The absolute contribution, the contribution of factors to the overall inequality, is the product of the elasticity and concentration index of each factor. The percentage contribution (the adjusted one) is calculated by dividing each absolute contribution to the overall concentration index. Whereas, the unadjusted percentage contributions can be obtained by diving the absolute contribution to the total explained portion of the concentration index (i.e., the overall concentration index minus the residual). When the absolute contribution has a similar sign to the overall concentration index, the percentage contribution will be positive, indicating that the factor has that much of a positive contribution to the inequality. This suggests that the overall inequality in the outcome variable would decrease by that percentage if the variable was equally distributed to each level of educational status and had no effect on outcome variable. The interpretation will reverse when the absolute contribution has an opposite sign to the overall concentration index. Results Background characteristics of study participants The median age of women was 26 years old with Inter Quartile Rang (IQR) of 22–31 years. Most of them were in the age group 25–34 years 744 (49.44%), married/in union 1,471 (97.81%), and from Oromia regional state 791 (52.58%). Regarding their educational status most of the women attended primary education 703 (46.71%) and only 121 (8.05%) women attended higher education. 1,107 (73.57%) women feel happy when they learned of pregnancy (Table 1). Table 1Background characteristics of study participants (n = 1,504).VariablesWeighted frequency (%)Exclusive breastfeeding statusp-valueYesNoResidenceUrban372 (24.77)1981740.049Rural1,132 (75.23)664468Age of the women15–24537 (35.70)3232140.31725–34744 (49.44)412332≥ 35223 (14.86)12796History of marriageOnly once1,326 (88.14)7585680.814Never and/or more than once178 (11.86)10474Educational status of womenNot attended school468 (31.11)2741940.043Primary education703 (46.71)408295Secondary education212 (14.13)12686Higher education121 (8.05)5467Partners educational status (n = 1,471)Not attended school452 (28.48)2711810.062Primary education624 (42.42)349275Secondary education263 (17.88)161102Higher education165 (11.22)8085Wealth status of the householdPoor594 (39.50)3472470.489Medium295 (19.59)175120Rich615 (40.91)340275ReligionProtestant426 (28.33)200226<0.001Orthodox515 (34.22)296219Muslim534 (35.49)350184Others29 (1.97)1613RegionAddis Ababa70 (4.69)3436<0.001Amhara329 (21.85)203126Oromia791 (52.58)492299SNNP314 (20.88)133181Feeling when learned of pregnancyUnhappy252 (16.76)1331190.377Mixed happy & unhappy145 (9.67)8758Happy1,107 (73.57)641466Wantedness status of the indexed pregnancyWanted then953 (63.34)5364170.423Wanted later468 (31.12)272196Not wanted at all83 (5.54)5330Place of deliveryHome586 (38.99)3422440.537Health facility918 (61.01)519399 Background characteristics of study participants (n = 1,504). Prevalence of EBF The overall prevalence of exclusive breastfeeding among Ethiopian women was 57.29 (95% CI: 54.79, 59.80%). The distribution of EBF was slightly higher among rural women: 198 (53.17%) and 664 (58.65%) of urban and rural women exclusively breastfeed their children, respectively. Education-related inequality in EBF practice of women The Wagstaff normalized concentration index for the education-related inequality in exclusive breastfeeding practice of women was −0.058 (95% CI: −0.113, −0.002, p-value = 0.039). Accordingly, there is a small but significant education gradient in exclusive breastfeeding practice. The negative sign reveals that EBF is disproportionately concentrated among the disadvantageous group (less educated women), meaning there is a pro-uneducated inequality in which women with low levels of education are more likely to practice EBF than the well-educated women. In order to get a distribution with an index value of zero (perfect equality), 4.35% (absolute value of the concentration index*75 (26)) of women who exclusively breastfeed would need to be redistributed from the less educated to the well-educated group of women. The concentration index is twice the area between the concentration curve and the line of equality (Figure 1). Concentration curve for the education-related inequality in exclusive breastfeeding practice among Ethiopian women. Decomposition of education-related inequality in EBF The Wagstaff decomposition analysis has been performed to decompose the overall education-related inequality in exclusive breastfeeding practices by variables and explain their contribution to the observed inequality. It quantifies how much of the observed education-related inequality was due to other variables. This was determined by calculating the coefficient, elasticity, concentration index, as well as the absolute and percent contribution for each factor. Given the pro-uneducated inequality observed in EBF practice in this study (indicated by a negative overall concentration index), if the independent variable is concentrated among the less educated group (negative concentration index) and exhibits a positive association with the outcome variable (positive coefficient), it would contribute positively (positive percentage contribution) to the overall inequality. Conversely, if the same variable shows a negative association with the outcome variable (negative coefficient), it would contribute negatively (negative percentage contribution) to the overall inequality. The reverse is true if the independent variable is concentrated among the well-educated group (positive concentration index). Accordingly, rural women are more concentrated among the less educated group (concindex = −0.504) and being rural increases the probability of practicing exclusive breastfeeding (coefficient = 0.009). Consequently, 18.8% of the pro-uneducated inequality in EBF practice was attributed by residence differences, meaning if residence was equally distributed across the ranking variable and had no effect on the exclusive breastfeeding practice, the observed education-related inequality would decrease by 18.8 percent. The wantedness status of the indexed pregnancy contributes the most (82.58%) to the pro-uneducated education-related inequality in exclusive breastfeeding practices among women. Additionally, more than one-third (29.31%) of the inequality is explained by geographical location (region). The act of giving birth at a health facility also explained −7.38% of the inequality, indicating that the total exclusive breastfeeding inequality would increase by 7.38% if place of delivery were equally distributed across the ranking variable and had no impact on exclusive breastfeeding practices. As indicated by the coefficient’s p-value, all other factors made a non-significant contribution to the overall concentration index. In general, in this study, the combination of variables fitted in the model explained the majority (85.21%) of the education-related inequalities in exclusive breastfeeding practices among women. This implies that there is 14.79% an unexplained difference in exclusive breastfeeding practices that cannot be accounted by the factors studied. Most of the determinants exhibit a low level of elasticity, meaning that their contribution patterns remain stable and unaffected by potential policy changes. However, the wantedness status of the indexed pregnancy stands out as the determinant that shows a different pattern (bigger elasticity and contribution), suggesting that it could have a more responsive impact on changes in the EBF practice of women (Table 2). Table 2The Wagstaff decomposition analysis results showing the contribution of various explanatory variables to education-related inequality in EBF practice among Ethiopian women.VariablesCoefficient (p-value)ElasticityConcentration indexContributionAbsolutePercentageResidenceUrban (Ref.)Rural0.009* (0.045)0.022−0.504−0.011 18.80 Age of the women15–240.032 (0.234)0.0430.1140.005−8.4125–340.006 (0.352)0.0120.1000.001−2.07≥35 (Ref.)RegionAddis Ababa (Ref.)Amhara0.105* (0.029)0.095−0.167−0.016 27.35 Oromia0.073* (0.032)0.105−0.011−0.001 1.96 SNNP−0.003 (0.113)−0.004−0.1060.0004−0.69History of marriageNever or above once (Ref.)Only once0.001 (0.456)0.0010.3540.0005−0.79ReligionProtestant (Ref.)Orthodox0.070 (0.066)0.1140.1330.015−26.01Muslim0.052 (0.085)0.072−0.111−0.00813.79Others0.041 (0.723)0.002−0.325−0.00071.15Wealth statusPoor0.017 (0.691)0.021−0.482−0.01017.49Medium−0.020 (0.662)−0.013−0.2550.003−5.57Rich (Ref.)Pace of deliveryHome (Ref.)Health facility0.003* (0.013)0.0090.4770.004 −7.38 Wantedness status of the indexed pregnancyWanted then0.105* (0.003)0.282−0.170−0.048 82.58 Wanted later0.091 (0.150)0.1020.1550.016−26.99Not wanted at all (Ref.)Total explained inequality−0.049885.21%Residual−0.008214.79%Overall inequality−0.058100%Coefficient = Marginal effects from the probit model; * = significant at p-value < 0.05; Ref. = Reference category. The bold values indicate significance at a p-value < 0.05 for independent variables only. The Wagstaff decomposition analysis results showing the contribution of various explanatory variables to education-related inequality in EBF practice among Ethiopian women. Coefficient = Marginal effects from the probit model; * = significant at p-value < 0.05; Ref. = Reference category. The bold values indicate significance at a p-value < 0.05 for independent variables only. Discussion The aim of the present study was to examine the inequality in exclusive breastfeeding practices among women in Ethiopia based on their education levels, using nationally representative data. The overall prevalence of exclusive breastfeeding (57.29%) was found to be comparable to previous studies conducted in Ethiopia, where exclusive breastfeeding rates were reported as 57.3% (14), 58% (11), 50.1% (27), and 60.9% (15). Our study revealed that exclusive breastfeeding is more prevalent among women with lower levels of education, indicating a pro-uneducated inequality. This disparity was quantified with a concentration index of −0.058, which is statistically significant at the 95% confidence level. This finding is consistent with previous evidence suggesting that EBF practices are often less favorable among more educated women (28–31). Several factors might explain this trend. Previous research has shown that women with higher education level are more likely to have demanding careers and shorter paid maternity leave. This professional pressure often necessitates returning to work before the infant reaches 6 months, thereby limiting the duration of exclusive breastfeeding (13, 16, 32, 33). In addition, higher educational attainment is often associated with increased body image concerns, which can negatively impact a woman’s commitment to exclusive breastfeeding. Traditional perceptions that breastfeeding might alter a mother’s body shape and adversely affect her health can exacerbate these concerns, potentially discouraging educated women from adhering to EBF practices (17–19, 34). Another possible explanation is that highly educated mothers often have higher household incomes, enabling them to afford infant formula. This is particularly relevant in urban areas where formula feeding may be perceived as a sign of modernity or higher social status. Limited access to comprehensive breastfeeding education and the influence of marketing strategies promoting formula as a convenient alternative can further contribute to this preference, potentially leading to reduced rates of exclusive breastfeeding among more educated women (35, 36). Contrastingly, other studies indicate that higher education levels often correlate with better knowledge of and adherence to breastfeeding guidelines. For instance, a global study reported that women with higher education levels were more likely to understand the benefits of EBF and thus were more likely to practice it (37). This is supported by the findings of similar study, which suggest that educated women have greater access to resources and support for breastfeeding (38). These studies argue that education equips women with the knowledge and skills necessary for successful breastfeeding, potentially leading to higher EBF rates. Therefore, improving EBF rates requires not only addressing educational disparities but also ensuring that all women receive comprehensive support to overcome barriers and promote successful breastfeeding practices. In the Wagstaff decomposition analysis, factors such as residence, region, place of birth, and wantedness status of the indexed pregnancy had a significant contribution to the observed education-related inequality in exclusive breastfeeding practice among women in Ethiopia. Women residing in rural areas are predominantly found within the less educated group of women, and they are more likely to exclusively breastfeed their children. Consequently, residence makes a notable positive contribution to the observed pro-uneducated inequality in EBF, indicating that women with lower levels of education are more prone to engaging in exclusive breastfeeding. The finding is in line with other studies done in Ethiopia (39–41) and Malaysia (42). This could be explained by the fact that urban women will have more diverse employment opportunities, reducing the time they can spend with their infants and potentially affecting exclusive breastfeeding practices. Alternatively, it could be attributed to urban mothers having greater access to alternative infant feeding options compared to their rural counterparts. Women from the Amhara and Oromia regions are disproportionately found in the less educated group of women and being from Amhara or Oromia is positively correlated with exclusive breastfeeding compared to women from Addis Ababa. Accordingly, approximately 29% of the education-related inequality in exclusive breastfeeding practices is attributable to this factor. Literature supports the assertion that women from larger cities, such as Addis Ababa, are less inclined to exclusively breastfeed their children for the recommended 6 months. A study from Addis Ababa indicated that only 44.2% of the mothers practiced EBF (43). Conversely, women who give birth at a health facility are predominantly present within the well-educated group of women and are inclined to engage in exclusive breastfeeding. Thus, it has a negative percentage contribution (decrease the inequality by some extent) to the observed education-related inequality favoring less educated women in exclusive breastfeeding practices among women in Ethiopia. As a result, encouraging educated women to deliver at health facilities is crucial, as it enhances their likelihood of practicing exclusive breastfeeding (44), thereby further reducing the pro-uneducated education-related inequality in EBF practice. The wantedness status of the indexed pregnancy made a significant, highly elastic, and big contribution to the observed pro-uneducated education-related inequality in exclusive breastfeeding practices of women in Ethiopia. Wanted pregnancies were mostly among women with lower levels of education, and these women were also more likely to breastfeed exclusively. It is supported by different studies (45, 46). Therefore, interventions should have to be designed to make all the pregnancies wanted among educated women to significantly reduce education-related inequality favoring less educated group in EBF practice of women. There are, of course, still limitations that should be considered when interpreting the results. Firstly, the reliance on self-reported data, particularly concerning exclusive breastfeeding practices, raises the possibility of recall bias or social desirability bias, where participants might overreport or underreport their behaviors. Furthermore, the cross-sectional design of the study restricts the ability to establish causal relationships, as it only captures associations without confirming the directionality of those relationships. Unmeasured confounding factors that were not included in the analysis due to data limitations of the PMA survey may also influence exclusive breastfeeding practices, potentially confounding the results. For instance, factors such as maternal mental health, employment status/occupation, workplace support, cultural beliefs, and healthcare access and quality were not considered in this study, yet they can significantly impact breastfeeding behaviors. Conclusion The proportion of exclusive breastfeeding practice among women in Ethiopia was relatively low. Furthermore, exclusive breastfeeding practice was disproportionately concentrated on the less educated group of women (pro-uneducated concentration). Thus, emphasis should be placed not only on educating women, but also on healthy habits that they can leave behind when they learn. The government should also address reasons, such as having a job that offer only 120 working days of paid maternity leave, which may prevent mothers from abandoning their exclusive breastfeeding habits when they learn. Residence, region, place of birth, and wantedness status of the indexed pregnancy made a significant contribution to the observed education-related inequality in exclusive breastfeeding practice among women in Ethiopia. The wantedness status of the indexed pregnancy made a more elastic contribution to the observed education-related inequality, implying that policy changes addressing pregnancy intention could readily reduce the observed inequality. Thus, the Ethiopian government is advised to design policies to ensure that every pregnancy is desired.