Engaging Fathers in Child Nutrition: Exploring Intra‐household Gender Dynamics in Nutrition Interventions in Ethiopia

ABSTRACT

This study aims to examine intra‐household gender dynamics in response to a nutrition intervention (maternal Behaviour Change Communication (BCC), paternal BCC, and food vouchers) aimed at improving IYCF practices using qualitative methods. Participants were drawn from a subset of households enrolled in a larger cluster – randomized controlled trial (RCT) conducted in rural Ethiopia. A total of 40 participants (20 mother–father pairs) from intervention and control households were interviewed separately to explore intra‐dyadic beliefs and household decision‐making. Furthermore, this study explores plausible mechanisms behind the main RCT finding that greater father involvement, while increasing knowledge, did not consistently improve infant and young child feeding (IYCF) outcomes. We find that BCC mothers had the tendency to adopt more gender‐equal beliefs, particularly regarding men's roles in childcare and household chores, whereas fathers were slower to shift their views. This may have led to increased discordance in gender norms within BCC households. By contrast, control couples often retained traditional views, but showed more intra‐couple alignment. Unexpectedly, control fathers were sometimes more progressive than control mothers, potentially due to higher education levels. We also find increased paternal engagement occasionally introduced conflicting priorities, as fathers asserted authority over household spending – sometimes at the expense of child‐focused nutrition. These findings underscore the complexity of engaging fathers in nutrition interventions and point to the need for strategies that address underlying gender norms and decision‐making dynamics within households.

Summary

• Mothers in BCC households became more gender‐equitable in their views, particularly regarding men's roles in childcare and domestic work, whereas fathers were slower to shift. This divergence increased discordance compared to control households, which tended to be more aligned, though typically traditional.

• Increased paternal involvement in household decisions sometimes generated conflicting priorities, as men asserted authority over resources and redirected food voucher savings away from child‐focused nutrition.

• These findings highlight the complexity of engaging fathers in nutrition interventions and underscore the importance of strategies that explicitly address gender norms and household decision‐making dynamics.

1. Introduction

Globally, 148 million children under the age of five suffer from stunting, defined as a height‐for‐age Z score below −2 SD, with 43% in Africa (World Health Organization 2023). Stunting, a key measure of chronic undernutrition, has profound short‐ and long‐term consequences on health, cognitive abilities, educational outcomes, and future income potential, creating a cycle of disadvantage across generations (Black et al. 2013; Hoddinott et al. 2013). In Ethiopia, 38% of children under five were stunted in 2016 (Ethiopia 2017). Inadequate infant and young child feeding (IYCF) practices, often due to insufficient food quantity and quality, as well as limited parental knowledge, are significant obstacles to improving nutrition (Alemayehu et al. 2015; Bilal et al. 2015). Increasingly, nutrition‐sensitive interventions are used to address these challenges (Escher et al. 2024; Santoso et al. 2021). These nutrition‐sensitive interventions typically target mothers (Manley et al. 2020). The logic is two‐fold. First, assuming that household decisions reflect some sort of bargaining process (Alderman et al. 1995), providing resources to mothers is seen as a way of increasing their decision‐making power (Gil‐García 2016). Second, it is argued that resources received by mothers are more likely to benefit children than those received by fathers (Basu and Maitra 2020; Ringdal and Sjursen 2021). Therefore, many programs focus on enhancing mothers' decision‐making power and control over resources, strengthening their influence on child health and growth.

However, these approaches neglect the potential role that fathers can play in ensuring children consume healthy diets. In many low‐income country settings, a large fraction of household income is generated by fathers and – in environments with large gender inequalities – they make most decisions on how household income is to be spent (Tommasi 2019). Furthermore, preparing nutritious food requires time; fathers' support with household chores is helpful.

Given this, interventions aimed at improving infant and young child diets are increasingly involving fathers (Martin et al. 2021; Morgan et al. 2017; Moura and Philippe 2023; Oryono et al. 2021). However, results are mixed. Some studies show positive associations between fathers' nutrition knowledge and child health outcomes, and fathers' direct involvement in feeding practices is associated with better diet diversity (Chary et al. 2011; Guerrero et al. 2016). Other studies show adverse effects when fathers are involved (Gazeaud et al. 2022; Han et al. 2022).

A cluster randomized controlled trial (RCT) in Ethiopia assessed the impact of involving fathers, through paternal Behaviour Change Communication (BCC) and food voucher program (Han et al. 2022). The study found that increased paternal involvement did not necessarily improve child feeding outcomes and may have reduced maternal decision‐making autonomy. A follow‐up study was conducted to provide a qualitative examination of differences in beliefs between BCC and control mothers, as well as between BCC and control fathers (Han et al. 2024).

In this qualitative investigation, we use interviews to explore behavioural, normative, and control beliefs around childcare, household labour, and decision‐making by group (BCC households vs. control households). Paternal BCC included sessions aimed at promoting gender‐equal beliefs for fathers. Although the maternal BCC program did not directly address gender equality, they featured role‐playing activities that illustrated gender‐equal decision‐making within couples. We hypothesized that couples in BCC households would exhibit greater concordance in gender‐equal beliefs compared to couples in control households, where more traditional views were expected to prevail. Specifically, we anticipated that both mothers and fathers in BCC households would adopt more progressive attitudes, with mothers and fathers aligning more closely on gender‐equal dynamics. In contrast, couples in control households were expected to maintain greater concordance in traditional beliefs due to limited exposure to gender‐equal messaging and peer influence. Furthermore, given that the main study found the group with increased father involvement showed the least improvement in child feeding practices (Han et al. 2022), we aimed to explore potential reasons why father involvement might negatively affect children's diets.

2. Methods

2.1. Study Context

We conducted a qualitative investigation of participant beliefs as part of a larger study, ‘What Promotes Healthy Eating? The Roles of Information, Affordability, Accessibility, Gender, and Peers on Food Consumption,’ carried out in Ejere, Ethiopia between 2017 and 2018. Ethical approval was granted by the institutional review board at Cornell University (USA, 1612006823), the Oromia State IRB (Ethiopia, BEIO/AHBHN/1‐8/2670), and Myungsung Medical College (Ethiopia).

The interventions were based on a Theory of Change that linked maternal BCC, paternal BCC, and food vouchers to changes in household food decision‐making and improvements in child diet diversity. The treatment groups were: maternal BCC only (T1); maternal and paternal BCC (T2); voucher only (T3); maternal BCC and food vouchers (T4); maternal BCC, food vouchers, and paternal BCC (T5); and a control group. Each BCC group consisted of 14 participants from the same village, meeting weekly with a trained facilitator for 1 h. Maternal BCC sessions lasted 16 weeks, covering topics such as types, diversity, quantity, preparation, and storage of complementary foods. Paternal BCC sessions lasted 12 weeks and included messages about diet diversity, malnutrition consequences in the first 2 years, fathers' roles in childcare, shared household labour, and gender‐equal decision‐making. Each household received a monthly food voucher worth 200 ETB (approximately US $10) for 4 months. Additional details are found in Han et al. (2021, 2022).

2.2. Qualitative Interview Design and Theoretical Considerations

We used Icek Ajzen's Theory of Planned Behaviour (TPB) for the development of semi structured qualitative interview guides and analysis of the agreement between couples (Ajzen 2011). The TPB suggests that behavioural intentions are shaped by three main types of beliefs: behavioural, normative, and control. Behavioural beliefs refer to a person's positive or negative evaluations of performing a particular behaviour. Normative beliefs involve the perceived social pressure to engage or not engage in a behaviour, influenced by cultural and societal norms. Control beliefs are related to an individual's perception of the difficulty or ease of performing the behaviour. These beliefs interact with each other and collectively influence the intention to perform a behaviour.

Identical questions were provided to both mothers and fathers, focusing on fathers' involvement in childcare, household chores, and decision‐making regarding household purchases. Fathers answered questions about their own behaviours, while mothers answered questions about their partners' behaviours. For instance, mothers were asked to report on the extent of fathers' involvement in household chores.

Questions were pilot tested, and three main topics were explored: fathers' roles in childcare, household chores, and household decision‐making. For each topic, we examined behavioural, normative, and control beliefs for mothers and fathers separately.

2.3. Data Collection

Subsets of participants from the trial were recruited for individual interviews using a maximum variation sampling strategy to capture heterogeneity. Variables from the quantitative data, including the knowledge score eligible child's age, residence, number of school years, and child dietary diversity score at baseline, were used for this purpose. Key variables for participant selection are shown in Table 1, while the characteristics of the selected participants for qualitative interviews from BCC and control households are presented in Table 2. The knowledge score refers to the proportion of correctly answered nutrition knowledge questions in the baseline survey. CDDS, an indicator of diet quality is calculated as the number of food groups consumed by the child in the previous 24 h, based on WHO IYCF guidelines. The asset score is a household wealth index constructed using principal component analysis (PCA) of household asset ownership variables.

Table 1.

Key variables used to select participants.

	Rural (N = 12)	Urban (N = 28)
Eligible child's age
< 12 months	3	4
12–15 months	2	6
16–20 months	1	2
CDDS score (FU)
2	3	2
3	1	5
4	1	3
5	1	1
6	0	3
Men's knowledge score
< 40	0	2
40–65	3	3
> 65	3	8
Men's school year
0	0	1
2–5	4	4
6–8	2	2
>= 9	0	7
Birth order
1	2	4
2	3	4
3	0	3
4	1	1
5	0	2

Note: CDDS = Child Dietary Diversity Score (range 0–7), based on the number of food groups consumed in the previous 24 h. Knowledge score refers to baseline paternal nutrition knowledge (proportion correct).

Table 2.

Participant characteristics.

	BCC households interviewed (N = 26)	Control households interviewed (N = 14)	All (N = 682) entire RCT sample from quantitative study
Maternal
Knowledge score	0.68 (0.14)	0.63 (0.14)	0.64 (0.11)
Married (%)	0.83 (0.38)	0.83 (0.4)	0.87 (0.33)
Age	27 (6.31)	27 (5.79)	28.37 (6.34)
Number of school years	6 (7.44)	2.66 (3.2)	4.43 (5.54)
Currently work (%)	0.5 (0.52)	0.16 (0.4)	0.54 (0.49)
Paternal
Knowledge score	0.61 (0.18)	0.64 (0.07)	0.6 (0.14)
Age	32.33 (6.44)	39 (15.81)	34.19 (8.13)
Number of school years	8.83 (7.9)	7 (3.34)	6.38 (6.47)
Currently work (%)	1 (0)	1 (0)	0.88 (0.31)
Child
Eligible child's age	12.66 (4.07)	12.5 (3.61)	12.34 (4.93)
Child's birth order	2.25 (1.05)	2.33 (1.5)	2.43 (1.49)
Household
Household size	4.66 (0.98)	4.33 (1.5)	4.7 (1.53)
Orthodox (%)	0.75 (0.45)	0.66 (0.51)	0.84 (0.35)
Oromo (%)	0.66 (0.49)	0.66 (0.51)	0.76 (0.42)
Rural (%)	0.41 (0.51)	0.16 (0.4)	0.44 (0.49)
Handwashing place near cooking area (%)	0.83 (0.38)	0.66 (0.51)	0.64 (0.47)
Asset score	0.11 (1.15)	0.82 (1.75)	0.09 (1.7)
IYCF practices
Child dietary diversity score	3.41 (1.62)	2.66 (1.03)	2.65 (1.64)
Minimum acceptable diet (%)	0.16 (0.38)	0.16 (0.38)	0.18 (0.38)
Minimum dietary diversity (%)	0.25 (0.45)	0.25 (0.45)	0.29 (0.45)
Minimum meal frequency (%)	0.66 (0.49)	0.66 (0.49)	0.54 (0.49)

Note: BCC and Control columns represent participants included in the qualitative interview study (N = 26 and N = 14, respectively). The ‘All’ column refers to the full sample from the quantitative RCT (N = 682) and is included for reference only. Values are means (standard deviations) for continuous variables and proportions (standard deviations) for binary variables. CDDS = Child Dietary Diversity Score. Knowledge score represents the proportion of correct answers on a nutrition knowledge module. The sample includes 26 participants from BCC households and 14 from control households.

Table 2 shows 41% of BCC participants and 16% of control participants resided in rural villages. Control households generally possessed more assets than BCC households. While fathers' educational attainment was similar between the groups, mothers in BCC households had more years of schooling compared to those in control households.

Participants were approached via kebele leaders or phone calls. Initially, twenty‐seven couples were approached, but seven were excluded from the qualitative sample: three because both parents were not interviewed, two because one parent refused to participate, and two because they missed the interview. Consequently, twenty couples (forty participants) from the maternal and paternal BCC group, the maternal and paternal BCC and food voucher group, and the control group were interviewed separately. We monitored the transcripts during data collection and found that no substantially new themes emerged after the eighteenth interview, suggesting that thematic saturation had been achieved. For the purpose of this study, we combined the maternal and paternal BCC group with the maternal and paternal BCC and food voucher group to compare the BCC and control groups.

Top‐performing enumerators with prior RCT survey experience were selected for the interviews. One male and one female interviewer, each with over 2 years of experience as enumerators but no prior qualitative interview experience, were trained by the principal investigator (PI) over a 14‐day period. They translated the interview scripts into Oromo and Amharic. Back translation was conducted post‐data collection and confirmed accuracy. Four pilot interviews during training were used for evaluation but not included in the data analysis. The PI and interviewers had no prior relationship with the participants. The PI observed the first eighteen interviews to ensure data quality, intervening when necessary to elicit further details.

2.4. Data Analysis

The audio recordings in Oromo and Amharic were transcribed and cross‐checked before being translated into English by native speakers of both languages. Interviews typically lasted about 40 min, with those involving the PI extending by approximately 15 min due to more in‐depth probing. The transcripts were reviewed to ensure all questions were addressed, with shorter durations generally due to less detailed probing.

The translated scripts were coded using Atlas.ti software, based on an initial coding framework created by the PI from pilot interview scripts. Thematic analysis was performed by three coders, including the PI, who organized the codes into categories reflecting common themes and patterns in the data. This process highlighted convergences and divergences between the BCC and control groups, revealing both similarities and differences. The themes were then simplified and categorized into gender‐equal or traditional beliefs. Gender‐equal beliefs were defined as those supporting gender‐equal dynamics within the household.

To ensure trustworthiness, we took multiple steps aligned with qualitative research standards. Credibility was strengthened through rigorous interviewer training, careful piloting of instruments, and direct oversight. Two bilingual enumerators (one male, one female), both with over 2 years of experience in quantitative survey work, were selected based on performance and trained intensively over a 14‐day period by the PI. During training, four pilot interviews were conducted to test question phrasing, sequencing, and probing strategies; these were used for feedback and refinement of the interview guide but were excluded from the final analysis. During data collection, the PI observed the first 18 interviews, occasionally intervening to enhance depth or clarify responses, ensuring consistency and quality. Participants were also allowed to select their preferred interviewer, further enhancing comfort and openness during the interview. Transferability was addressed by applying a maximum variation sampling strategy, deliberately selecting participants with diverse child ages, household asset levels, locations (rural and urban), and education levels. This approach, drawn from the quantitative survey data, aimed to capture a wide range of household dynamics relevant to intra‐couple gender norms and decision‐making practices. Dependability was ensured through a transparent and replicable analytic process. All interviews were transcribed, translated, and verified for accuracy. Three coders, including the PI and two additional trained researchers, independently applied a shared coding framework developed from pilot interviews and aligned with the Theory of Planned Behaviour. Coding discrepancies were reconciled through discussion, and regular team debriefings were held to refine themes and ensure analytic consistency.

For transparency and interpretability, responses were grouped into ‘gender‐equal’ or ‘traditional’ belief categories based on their implications for intra‐household dynamics. This categorization enabled us to identify broad patterns – such as the proportion of participants expressing gender‐equal beliefs – summarized in Tables 3 and 4. However, no inferential statistics were conducted, as the study was neither powered nor designed for hypothesis testing. These percentages are therefore intended to illustrate descriptive tendencies within this purposively selected sample. Importantly, these coded summaries function as entry points for cross‐group comparison, while the primary analysis was thematic, grounded in inductive coding of verbatim transcripts. This approach allowed us to preserve and interpret the richness and complexity of participants' lived experiences, using the categorical summaries as a scaffold for identifying concordance and discordance across households.

Table 3.

Gender‐equal mother and father – comparing couples who are in BCC and control households.

Constructs	Question	BCC mother gender‐equal	BCC father gender‐equal	Difference BCC households	Control mother gender‐equal	Control father gender‐equal	Difference control households
	Men's role in childcare
Behavioural	If children are younger than two years old, fathers have a limited role to play.	64	64	0	43	71	−29
Behavioural	Tell me about what men can do to keep the child happy and healthy.	93	86	7	86	86	0
Behavioural	Why are you (your husband) not involved in these activities?	64	64	0	57	71	−14
Normative	It is common for men to be involved in childcare activities.	79	50	29	57	43	14
Normative	It is culturally acceptable for men to do childcare activities.	71	57	14	71	71	0
Control	It is difficult for men to do childcare activities.	64	64	0	71	71	0
	Men's role in household chores
Behavioural	Why are you (your husband) not involved in these activities?	36	21	14	0	43	−43
Normative	It is common for men to be involved in household chores.	64	36	29	43	43	0
Normative	It is culturally acceptable for men to do household chores.	79	29	50	29	43	−14
Control	It is difficult for men to do household chores.	57	36	21	14	43	−29
	Intra‐household decision‐making
Behavioural	Men make better decisions than women.	57	64	−7	57	43	14
Behavioural	What do you feel when men give women power to make household decisions?	64	71	−7	43	71	−29
Normative	What would others say if men gave women power to make household decisions?	43	29	14	14	43	−29

Note: The responses were grouped into ‘gender‐equal’ or ‘traditional’ belief categories based on their implications for intra‐household dynamics. The results are presented as the percentage of BCC and control women and men responding in a gender‐equal manner, highlighting overall tendencies in the data. The ‘Difference’ column indicates the proportion of households in disagreement due to differences in beliefs between couples (gender‐equal vs. traditional). A positive difference reflects disagreement driven by mothers holding gender‐equal beliefs, while a negative difference indicates disagreement due to fathers adhering to gender‐equal beliefs.

Table 4.

Concordance in gender‐equal direction – comparing couples in BCC and control households.

		BCC households		Control households		Difference: BCC ‐ Control
Construct	Question	Discordant: mother gender‐equal	Concordant Gender‐equal	Discordant: mother gender‐equal	Concordant Gender‐equal	Discordant: mother gender‐equal	Concordant Gender‐equal
	Men's role in childcare
Behavioural	If children are younger than two years old, fathers have a limited role to play.	14	50	0	43	14	7
Behavioural	Tell me about what men can do to keep the child happy and healthy.	14	79	0	86	14	−7
Behavioural	Why are you (your husband) not involved in these activities?	21	43	0	57	21	−14
Normative	It is common for men to be involved in childcare activities.	29	50	14	43	14	7
Normative	It is culturally acceptable for men to do childcare activities.	29	43	14	57	14	−14
Control	It is difficult for men to do childcare activities.	21	43	0	71	21	−29
	Men's role in household chores
Behavioural	Why are you (your husband) not involved in these activities?	36	0	0	0	36	0
Normative	It is common for men to be involved in household chores.	36	29	29	14	7	14
Normative	It is culturally acceptable for men to do household chores.	50	29	14	14	36	14
Control	It is difficult for men to do household chores.	43	14	0	14	43	0
	Intra‐household decision‐making
Behavioural	Men make better decisions than women.	29	29	29	29	0	0
Behavioural	What do you think is the benefit when men give women power to make decisions on household purchases?	14	50	14	29	0	21
Normative	What would others say if men gave women power to make household decisions?	21	21	14	0	7	21

Note: The responses were grouped into ‘gender‐equal’ or ‘traditional’ belief categories based on their implications for intra‐household dynamics. The results are presented as the percentage of BCC and control households exhibiting concordance or discordance, highlighting overall tendencies in the data. The ‘Difference’ columns compare BCC and control households in terms of concordance and discordance. A positive difference in discordance reflects more BCC households in disagreement due to mothers holding gender‐equal beliefs compared to control households. Similarly, a positive difference in concordance (gender‐equal) indicates more BCC households in agreement in a gender‐equal direction.

We define concordance as the agreement in responses between couples. We classify concordance into two types: (1) gender‐equal concordance, where couples express gender‐equal beliefs, and (2) discordance, where the mother expresses gender‐equal beliefs while the father holds traditional beliefs. Additionally, we investigate how increased father involvement may have adversely affected children's diets through exploration through the lens of household dynamics.

2.5. Use of AI

No use was made of Artificial Intelligence Generated Content (AIGC) in developing any part of this manuscript.

3. Results

Tables 3 and 4 summarize gender‐equal beliefs and household dynamics in BCC and control households. Table 3 highlights the differences in gender‐equal beliefs between mothers and fathers, while Table 4 presents intra‐dyadic comparisons within couples, exploring how mothers and fathers in the same household share or differ in their beliefs and practices, examining the nature of discordance and concordance.

While we find concordance in the gender‐equal direction was similar between BCC and control households in some constructs, BCC households showed greater discordance due to mothers in BCC households tending to adopt more gender‐equal beliefs, while fathers retained more traditional views. In contrast, control households generally had less discordance, as both mothers and fathers held more traditional beliefs.

3.1. Men's Role in Childcare

Table 3 reveals differences between mothers and fathers in their gender‐equal beliefs about men's roles in childcare. Greater discordance in BCC households was observed primarily in the normative constructs. In BCC households, 79% of mothers believed men's involvement in childcare was common, compared to only 50% of fathers. Similarly, 71% of mothers believed men's involvement was culturally acceptable, compared to 57% of fathers. Table 4 shows that this divergence translated into 29% of BCC households experiencing discordance due to mothers holding more gender‐equal views, compared to 14% in control households. However, BCC households also achieved greater concordance in the gender‐equal direction (50%) compared to control households (43%).

Despite these differences, both BCC and control households agreed that fathers' primary role in childcare was financial, such as earning money to purchase food and cover health‐related expenses. Fathers' involvement in direct childcare, such as holding or feeding the child, was often seen as limited to occasions when mothers were busy with other household chores. About half of the mothers and fathers in both groups viewed men's childcare roles as socially acceptable, but fathers' work routines were cited as a reason for their limited involvement.

Our commitment to work from outside and bring home for proper growing of our child is very important

(BCC father)

Since women has nurturing responsibility. Father also has responsibility to nurture. Since father has other way of nurturing and no ability to take care of the small children its mothers responsibility.

(Control mother)

3.2. Men's Role in Household Chores

Differences in gender‐equal beliefs between mothers and fathers were most pronounced in the normative constructs related to household chores. Table 3 shows that in BCC households, 79% of mothers believed men's participation in household chores was culturally acceptable, compared to only 29% of fathers. Similarly, 64% of BCC mothers viewed men's involvement in chores as common, compared to 36% of fathers. In contrast, control households exhibited smaller differences: 29% of mothers and 43% of fathers believed men's involvement was culturally acceptable, resulting in fathers being slightly more gender‐equal than mothers. On the behavioural construct‐ specifically, the question about why men are not involved in these activities‐control households revealed a notable difference, with 43% of fathers expressing gender‐equal beliefs compared to 0% of mothers.

From Table 4, the discordance in BCC households regarding household chores was primarily driven by mothers adopting more progressive beliefs about men's roles. For instance, 36% of BCC households displayed discordance on the behavioural construct question ‘Why are you (your husband) not involved in these activities?’ due to mothers being more gender‐equal, compared to 0% discordance in control households. These findings suggest that while BCC successfully shifted maternal beliefs toward gender equality, they also may have created tensions as fathers were slower to adapt to these progressive norms. Control households exhibited less discordance but retained traditional roles, as reflected in mothers reinforcing the conventional division of labour.

If the mother cleans the clothes or cooks injera; she does for the family or for the child. It is good if that Bahu (husband) supports her by feeding the child while she washes clothes. Therefore, he (husband) is not expected to leave everything to her.

(BCC mother)

Since they do not know how to work (on household chores), most of the time it is the work that matters to females. I think it (household chores) concerns them (fathers), but I think is not familiar to them (fathers).

(Control mother)

3.3. Intra‐household Decision‐Making

Table 3 shows BCC households with relatively small differences between mothers and fathers. For instance, 64% of mothers and 71% of fathers believed it was beneficial for men to give women power to make household decisions. In control households, this difference was larger, with 43% of mothers and 71% of fathers agreeing, creating a 29% difference and suggesting that control fathers were more progressive in this belief than control mothers. Similarly, for the question about what others might think if men gave women decision‐making power, 43% of mothers and 29% of fathers in BCC households agreed in gender‐equal direction, while 14% of mothers and 43% of fathers agreed in gender‐equal direction, again reflecting more progressive beliefs among fathers in control households.

From Table 4, we find discordance due to mothers' gender‐equal belief on decision‐making was similar in both BCC and control households. However, BCC households achieved higher concordance in the gender‐equal direction for the question ‘What do you think is the benefit when men give women power to make household decisions?’ with 50% concordance in BCC households compared to 14% in control households. On the question ‘What would others say if men gave women power to make household decisions?’ 21% of BCC households and 0% of control households showed concordance in gender‐equal direction.

He makes the big decisions. If he does not accept my decision, I will say nothing.

(BCC mother)

Men make better decisions compared to women, and their decisions are more acceptable, though the difference may be very small.

(Control father)

3.4. Exploration Into Possible Channels of Father Involvement Negatively Affecting Children's Diet

Findings from the quantitative study indicate that the BCC program improved nutrition knowledge for both mothers and fathers. The greatest improvements in IYCF practices were observed when maternal BCC was combined with a food voucher. However, when paternal BCC was added to maternal BCC and the food voucher, there was an increase in father involvement but a smaller improvement in IYCF practices compared to maternal BCC and voucher alone. This suggests that the paternal BCC program, while enhancing father involvement, may have inadvertently had a negative impact on children's diets. Figure 1 outlines the potential steps in this process, explored further through qualitative investigation.

Figure 1.

Proposed channels of father involvement negatively affecting children's diet.

3.5. BCC Empowering Mothers and Fathers

The paternal BCC program aimed to empower both mothers and fathers by increasing their knowledge of child nutrition and encouraging them to take active roles in household decision‐making. For mothers, the program boosted confidence and decision‐making power. Many BCC mothers, empowered by their increased knowledge, began negotiating with fathers on food purchases and making decisions to improve children's diets.

My husband is more knowledgeable than me, and I may also know more than my husband. So, not only can men decide better but also women could know better than men and men could know better than women.

(BCC mother)

This shift was significant, as many control mothers viewed decision‐making as a burdensome responsibility due to their lack of confidence or perceived knowledge, with one control mother stating,

That is a matter of competence. There are women having more knowledge than men. And there are also women who are less competent than their husbands. So, it is based on her knowledge and competence.

(Control mother)

For fathers, the program provided a sense of empowerment by increasing their perceived knowledge about nutrition and household decision‐making. Fathers believed that being knowledgeable gave them the right to lead decisions, particularly in areas of food purchases.

It is a matter of competence. If I have the knowledge, I should make the decisions, but if my wife knows better, I will let her decide.

(BCC father)

This dual empowerment created both opportunities and tensions as both parents sought to exert influence over food purchases, potentially leading to disagreements.

3.6. Paternal BCC Increasing Father Involvement in Food Purchase Decisions

The paternal BCC program significantly influenced fathers' involvement in household food purchase decisions. Fathers began to perceive themselves as actively participating in decisions, which traditionally were handled by mothers, except for the purchase of meat. However, the meaning of ‘father involvement’ varied. For some fathers, it meant guiding or advising mothers based on their newfound knowledge, while for others, it involved directly imposing their preferences on household purchases.

The father can keep the child happy through educating and buying good clothes. The child can also be happy if the father advises the mother and promises to buy good clothes.

(BCC father)

This involvement, however, revealed key differences in preferences between mothers and fathers. Mothers, empowered by the maternal BCC program, preferred to use food vouchers to buy nutrient‐rich foods like butter, fruits, and vegetables, emphasizing the importance of improving the child's diet.

I withdraw from my (bank) account to buy butter, fruits, and vegetables. I purchase what I need with my own money.

(BCC mother)

On the other hand, fathers often preferred to use the food voucher as a way to save money, reallocating the household budget for nonfood expenses or future needs.

We use it (the money we save by receiving voucher) for different things for home and also buy clothes.

(BCC father)

3.7. Dynamics of Disagreement: Negotiation and Unequal Decision‐Making Power

Disagreements between mothers and fathers over food purchases revealed unequal decision‐making power. While the paternal BCC program aimed to promote gender‐equal decision‐making, many BCC fathers retained traditional authority, which often dominated discussions. BCC mothers, despite their empowerment, faced challenges in negotiating effectively when disagreements arose. Similarly, a control mother emphasized the traditional hierarchy in decision‐making.

He makes the big decisions. If he does not accept my decision, I will say nothing.

(BCC mother)

He is superior in decision (compared to me) and he decides.

(Control mother)

Fathers, often framed decision‐making as collaborative but ultimately retained control. The increase in father involvement, while intended to promote joint decision‐making, often led to fathers asserting greater authority, thereby limiting mothers' ability to implement decisions aligned with maternal BCC goals.

I am the one who brings the idea and I discuss with her. After the discussion, I am the one who first make a decision. I am the first one to make the decision that it must be purchased and bring the idea.

(Control father)

3.8. Influence on Food Voucher Use

The increased involvement of fathers in food purchase decisions also influenced how food vouchers were used. In many cases, vouchers were not used to supplement the household food budget but instead substituted for previous food spending, allowing saved money to be redirected to nonfood expenses. About one‐third of BCC fathers expressed satisfaction with this approach. Conversely, more mothers viewed food vouchers as an opportunity to buy additional food, improving household nutrition. This disconnect between fathers' and mothers' preferences further highlighted the tension created by increased father involvement and its potential to crowd out the positive impacts of maternal BCC.

The food voucher helps us reduce our spending on food purchases.

(BCC father)

We discuss and agree on the amount of money to be spent on food purchase and the money to be saved. When we receive 200 birr (as food voucher), we use additional 100 birr for food purchase using our own money and we save 100 birr.

(BCC father)

Voucher helps us save our income that we get from other different activities such as salary. We use our salary to buy and feed the child before receiving voucher. But after receiving voucher, we are saving our salary for different things and for the future. We are saving it for the future, and we are keeping it on our account, we can use it for different things.

(BCC father)

4. Discussion

In BCC households, while mothers showed significant shifts toward gender‐equal beliefs, fathers were slower to adapt, leading to increased discordance within couples, particularly in normative beliefs around men's roles in childcare and household chores. Conversely, control households exhibited less discordance, with both mothers and fathers generally maintaining traditional beliefs. Control fathers were sometimes more progressive than control mothers in specific domains, such as decision‐making. We speculate that this may be due to longer years of schooling among fathers in the control areas. This highlights the influence of the BCC program in shifting maternal beliefs but also underscores the challenge of achieving alignment between parents.

The study also explored why increased father involvement in the intervention group did not lead to the expected improvements in child feeding practices in the quantitative study. While the BCC program aimed to empower both parents, the findings suggest that fathers' increased participation introduced unintended tensions in household decision‐making. The BCC program led fathers to take a more active role in household, but this often led to competing priorities. Fathers frequently prioritized saving or reallocating food voucher savings toward nonfood expenses, conflicting with mothers' focus on enhancing children's diets. Despite efforts to promote joint decision‐making, many fathers retained traditional authority, framing decisions as collaborative but ultimately asserting control. Although the program increased both parents' confidence in decision‐making, this dual empowerment created tensions, with fathers often retaining greater decision‐making power and ultimately making final decisions that were unfavourable to improving children's diets. These dynamics influenced food voucher use, with many fathers reallocating saved funds to nonfood expenses, diluting the program's nutritional impact. Social norms also limited the translation of gender‐equal messages into household practices, as fathers often supported shared decision‐making in theory but maintained authority in practice.

These findings are consistent with results from other settings. Systematic reviews of BCC and mHealth interventions in LMICs highlight both opportunities and risks in engaging men (Lwamba et al. 2022; Cheng et al. 2020; Kirkwood et al. 2022). While such programs can improve communication and health‐related knowledge between couples, studies suggest they may also inadvertently reinforce men's roles as household gatekeepers – particularly when interventions do not explicitly address gendered access, control, and digital literacy (Lwamba et al. 2022; Jennings and Gagliardi 2013). In some cases, providing men with direct access to health information increased their involvement in family health decisions but led to unilateral decision‐making rather than joint deliberation, which generated tension between couples (Jennings and Gagliardi 2013). These findings underscore the importance of adopting a gender‐transformative approach in the design of health programs, ensuring that male engagement enhances, rather than displaces, women's agency and promotes equitable decision‐making. Without such safeguards, interventions risk entrenching existing power imbalances or introducing new sources of intra‐household conflict.

This study has several strengths, including the use of a theoretically grounded and pilot‐tested interview guide, that distinguished behavioural, normative, and control elements, as well as maximum variation sampling to capture a diverse range of participant experiences. To minimize unconscious bias, responses were systematically coded, allowing for the clear identification of outlier cases. However, caution is warranted when interpreting the tables, as the study was not initially designed for quantitative analysis and lacked sufficient statistical power for some of the comparisons made.

This study is subject to several limitations. First, a greater proportion of participants in the BCC group were drawn from rural areas, where traditional gender norms are generally stronger. This imbalance may have introduced a downward bias, potentially making BCC households appear less gender‐equitable than they might have been. That said, the relatively higher educational attainment of BCC mothers may have partially mitigated this bias. Second, while enumerators had substantial experience conducting RCT surveys, they lacked formal training in qualitative research. To address this, we implemented a multi‐step quality assurance process that included the selection of top‐performing enumerators, 14 days of intensive training, pilot interviews, and structured oversight by the PI for a subset of interviews to ensure consistency and depth of data collection.

5. Conclusion

The paternal BCC program aimed to empower both parents, but its effects were multifaceted and sometimes worked in opposite directions. While it increased father involvement and knowledge, it also shifted household decision‐making dynamics, often reinforcing fathers' control and limiting mothers' autonomy in food purchase decisions. This imbalance, combined with differences in food preferences and negotiation dynamics, may have diluted the positive effects of maternal BCC and food voucher on IYCF practices. These findings illustrate the challenges of involving fathers in nutrition interventions and highlight the importance of creating more effective approaches to enhance their participation in improving children's diets.

Author Contributions

Yaeeun Han, David Pelletier, and John Hoddinott conceptualized and designed the research study. Yaeeun Han and Jieun Kim performed data collection. Yaeeun Han and David Pelletier analysed the data. Yaeeun Han wrote the first draft. Yaeeun Han, David Pelletier, and John Hoddinott reviewed and edited the manuscript. All authors read and approved the final manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data are available from the authors on request.