Japan introduced paternity leave legislation as early as 1991, allowing fathers to take up to 12 months of paid leave before their child's first birthday (5). Despite this provision, the uptake of paternity leave among Japanese men remained significantly low until 2017, when political reforms were introduced to provide more flexible paternity leave (5). Currently, Japan offers the second-longest total duration of paternity leave among the Organisation for Economic Co-operation and Development (OECD) countries, however uptake remains low, with only 14% of eligible fathers utilizing the benefit in 2021 (6). In contrast, countries like Sweden and Finland reported moderate to high rates of paternity leave uptake, reaching up to 80% in 2021. Paternity leave legislations of these countries share common features such as relatively long leave durations, high financial allowances, and high flexibility, encouraging greater participation (7). Similarly, countries like Spain and Estonia have experienced a rapid increase in paternity leave uptake, reaching around 70% in 2021, following political reforms that extended leave durations and improved financial support (6).

The low uptake of paternity leave in Japan, despite legislative reforms, suggests the presence of persistent barriers. These barriers can be examined using the social-ecological model, which highlights multiple levels of determinants on behavior. These levels include the individual level, with factors such as age, education, and income; the relationship level, involving peers, partners, and family; the community or organizational level, such as schools or workplaces; and societal factors, including political and cultural determinants (8). Previous studies in Europe and the United States have applied the social-ecological model to identify key factors influencing fathers’ use of paternity leave at different levels (9–12). At the individual level, employment and financial stability (10, 11) as well as prioritization of childcare responsibilities (12) are significant determinants. At the relationship level, the employment status of a partner also plays a critical role (10). Organizational and cultural barriers are particularly prominent in male-dominated professions, such as law, where workplace policies and support are crucial for fathers to take paternity leave (9, 11).

Despite extensive research in western countries, studies examining fathers’ perceptions of paternity leave in Japan remain limited. This gap may also be due to gender norms that disproportionately place the childcare burden on women, and Japan's collectivist work culture, which often discourages deviations from traditional norms (13). A brief pilot survey conducted by an international research team in 2024 (14) and a study conducted in 2017 highlighted that cultural norms and workplace barriers were primary reasons why fathers chose not to take paternity leave (15). However, these studies did not comprehensively address determinants of fathers’ perception of paternity leave across different levels as outlined in the social-ecological model. To fill this gap in the literature, our study intends to explore fathers’ perceptions of the factors associated with taking paternity leave in Japan, by applying the social-ecological model.

We conducted 21 in-depth online interviews with fathers between September 18, 2021 and October 23, 2021 to obtain information on their perceptions of paternity leave. A sample size of approximately 15 participants is deemed sufficient for qualitative analysis of interview data addressing a specific research question (16). The data collection was stopped upon the attainment of data saturation, defined as a point at which further interviews no longer generated new information (17). We also conducted key informant interviews with four supervisors at companies between September 30, 2021 and October 22, 2021 to understand their opinions on paternity leaves and ensure that the implications of this study are relevant at workplaces. The interviews with company supervisors were designed as key informant interviews with a smaller sample size and were analyzed separately from the more extensive interviews conducted with fathers (18). However, in the discussion section, findings from the fathers’ interviews were integrated with insights from the company supervisors to ensure that the implications of this study are meaningful and relevant in workplace settings. The first author (NM) contacted the initial participants (18 fathers and two company supervisors) through a personal social network. The remaining participants were contacted through early participant referral by four of the initial participants. We contacted participants via social network services or email.

Interviews were conducted in Japanese, and we used separate semi-structured interview guides for 21 fathers and four company supervisors (Supplementary Material). Our interview guide for fathers included questions on their age, occupation, education, children, and experiences and perceptions of paternity leave. Our interview guide for company supervisors included questions on age, occupation, education, children, and opinions about paternity leave. All participants agreed to video-record the interview (except for one company supervisor, who participated only by voice because a video camera was not available).

The Research Ethics Committee of Tokyo Medical and Dental University (Institute of Science Tokyo) approved this study (M2021-128). All the participants provided written informed consent. Each interview lasted for approximately 60 min. Before transcribing the data, the participants were anonymised to protect their privacy. The first author transcribed the interviews verbatim and typed the notes in Japanese.

Reflexive thematic analysis (TA) was adapted to analyze the data from the fathers and supervisors at companies (19–21). In reflexive TA, themes are generated “at the intersection of data, analytic process and subjectivity” (19–21). It integrates the research question, the authors’ knowledge, perspectives, and experiences, alongside the content of the data, to generate themes with validation (21). This method was selected because the first author was the father of a one-year-old infant and a regular employee of a private company. These factors made reflexive TA suitable, as it enabled a deeper exploration of the participants’ experiences. The final theme list was refined through an iterative process of merging and revising the codes, facilitated by discussions between NN and NM. We followed the Standards for Reporting Qualitative Research (22) in reporting this study. The first author performed thematic coding and developed a codebook in English with the second author (NN). Deductive coding was conducted based on a previous study on paternity leave (10), followed by inductive coding and modification of the initial codes. The final codes were organised based on the social-ecological model (8). ATLAS.ti (version 9) was used for data analysis.

We considered reflexivity throughout data collection and analyses (23). The first author's personal circumstances—being a student, the father of a one-year-old infant, and a full-time employee at a private company—may have influenced the research process. Living in an urban area with a spouse and child as a nuclear family without co-residing grandparents, the first author's family did not rely on multigenerational support, and the child was in good health. These factors likely influenced the study in three key areas: research design, data collection, and analysis. In terms of research design, the absence of co-residing grandparents may have limited the first author's awareness of how family-based social support, including that provided by grandparents, impacts parental leave decisions. Similarly, as a salaried employee in a corporate environment, the focus may have leaned toward workplace policies, potentially overlooking the unique challenges faced by self-employed or informal workers. Nevertheless, the study design, coinciding with the first author's own eligibility for childcare leave, may have brought nuanced insights into contemporary Japanese parental leave policies. For data collection, conducting interviews when the first author's child was over one year old may have shifted the focus toward post-birth parenting experiences, with less emphasis on pre-birth decisions. However, the first author's firsthand experience likely facilitated deeper and more empathetic engagement with participants, enriching the inquiry into parental leave practices. Finally, regarding analysis, the first author's active parenting responsibilities may have heightened sensitivity to themes related to childcare and parenting, while potentially underemphasizing broader aspects such as workplace advancement or personal leisure.

To separate the experiences of the author from those of the participants and minimize interpretation, we asked them additional questions whenever information was omitted from their answers. Furthermore, we explained that the participants could freely state their opinions because the interview data were anonymised before transcription. This approach was employed to ensure that the participants avoided criticism from family members or employers. Additionally, we attempted to minimise our interpretation during coding.

Table 1 outlines the characteristics of the 21 fathers. Most (71%) had only children. Nearly half (48%) of the participants were postgraduate students. Slightly below half (43%) had finished or continued to attain a paternity leave of six days or more.

We organised the themes into four levels of the social-ecological model: individual, relationship, community, and societal (refer to Table 2) (8).

Fathers’ career aspirations: Some fathers mentioned career aspirations as a theme associated with attainment or length of paternity leave. They perceived that paternity leave would affect their promotion, continuation of their current occupational roles, and competition with colleagues. A respondent noted a barrier to taking paternity leave,

In contrast, a father referred to the positive aspect of paternity leave for career development by stating that childcare can support the career by providing experiences that fathers do not get from the work though he had no experience of taking paternity leave.

Fathers’ commitment to supporting infants or mothers. Most fathers suggested that their individual commitment to supporting infants or mothers was a reason for taking paternity leave. They had taken paternity leave because of their ideological commitment as fathers to raise infants or as husbands to support their partners. Regarding commitment to raising infants, one father noted,

As for support for mothers, one father said,

Mothers’ or infants’ need for fathers’ involvement. Almost all fathers reported that they considered taking paternity leave in response to the mothers’ support needs, requests, or infants’ care needs. Regarding mothers’ support needs, fathers noted that their partners required physical or mental support after giving birth, which encouraged them to take the paternity leave.

Some fathers said that their wives requested them to take the paternity leave, or that their needs motivated them to do so.

Table 3 presents the characteristics of the company supervisors who participated. All were regular employees. The supervisors mentioned themes related to paternity leave from the broader perspective of long-term effects and impact on the organisation. First, all of them pointed out that the experience of taking paternity leave helps fathers improve work productivity after paternity leave. A company supervisor noted,

Another supervisor noted,

Second, some company supervisors mentioned that paternity leave could improve mothers’ productivity. One of them noted,

Another company supervisor stated,

Third, one company supervisor mentioned that paternity leave attracts colleagues’ attention to their work-life balance. He stated,

Regardless of the company supervisors’ perceptions, they did not proactively encourage fathers to take paternity leave. One of them said,

This study had several limitations. First, the first author, as a full-time employee at a private company and a postgraduate student, recruited participants primarily through personal networks. As a result, nearly half of the fathers were postgraduates, and most were regular employees. This recruitment approach may have not included perspectives from fathers with different occupational backgrounds, such as self-employed individuals, farmers, or freelancers, who may face unique challenges with childcare leave, including income loss or job insecurity. Furthermore, fathers with potentially higher childcare leave needs, such as those raising children with congenital conditions, may have also been underrepresented in the sample. Therefore, the results may not be transferable to fathers with characteristics substantially different from those of the study participants. Likewise, the new findings—the promoting influence of fathers’ career aspirations, mothers’ or infants’ need for fathers’ involvement, and fathers’ work location—may not be transferable to countries with low inclusiveness (39), wide economic gender gap (40), or poor digital adoption (41), respectively. Second, another limitation of the study was the small number of company supervisors included as key informants in the interviews (18). Third, the participating fathers might have avoided sensitive topics, such as medical history, sexual problems, or financial difficulties. However, these topics can fall into themes such as the mothers’ or infants’ need for fathers’ involvement or the paternity leave policies.