Primary Outcomes (explanation)
More specifically, we have a few different ways of measuring the establishment of a space.
1) Decision to establish. This will be measured at the end of August 2024, and can take two forms: schools can either decide to design their own lactation space, or can decide to purchase a mobile unit from our partner NGO. We measure this through the administrative records of our partner NGOs, who ask the schools to fill out forms in either case.
2) Mode of establishment. We will define two separate variables to distinguish between the design of an own lactation space and the decision to purchase a mobile unit from our partner NGO.
3) Effective establishment. This will be measured in endline visits or trough photographic proof of establishment.
We will also measure schools’ adoption of lactation policies and related practices. First, we ask the headteacher whether the school has a lactation policy. Second, we ask teachers whether they are oriented, when they start working at the school, on which space they can use for breastfeeding or expressing milk and on the rules about breastfeeding and expressing milk during working hours. Third, we also ask them whether there is an agreement between mothers and the head teacher regarding break time to breastfeed or express milk.
Teacher-level outcomes will be measured through teacher surveys administered at baseline and endline.
Primary teacher outcomes include the following families:
1. Demand, attitudes, and beliefs towards breastfeeding-related policies
For demand, we will capture teachers’ revealed and stated demand for breastfeeding-related workplace accommodations. Measures include willingness to contribute resources toward the establishment of lactation spaces, expressed support for workplace breastfeeding amenities, and prioritization of breastfeeding-related investments relative to other school expenditures.
We will construct a standardized demand index combining survey items related to willingness to financially contribute to lactation spaces; support for allocating school resources toward breastfeeding accommodations; stated prioritization of lactation spaces relative to alternative expenditures; and perceived importance of breastfeeding-related workplace policies. Higher values correspond to greater demand for breastfeeding-related workplace accommodations.
For attitudes, we will measure teachers’ attitudes regarding breastfeeding-friendly schools. Survey items capture agreement with statements regarding the respondent’s support for the establishment of lactation spaces. Higher values correspond to more favorable attitudes toward breastfeeding-related accommodations.
For beliefs, we will capture teachers’ beliefs about the social and professional consequences of breastfeeding accommodations in schools. Survey items measure perceptions regarding beneficiaries’ and their spouses’ reactions, expected effects of lactation rooms and breastfeeding policies on teachers’ breastfeeding at school, absenteeism, motivation or performance, fertility, and willingness to work in such schools, and concerns about breastfeeding in professional settings, including reservations about using lactation spaces and workplace taboos around discussing breastfeeding. Items will be coded such that higher values correspond to more supportive beliefs.
2. This outcome family measures broader demand for, attitudes toward, and perceptions of family-friendly work environments.
For demand, we will capture the importance that teachers give to flexible working arrangements and/or crèches at the workplace, as well as their likelihood of mentioning other family-friendly workplace policies as priorities for investment in the school.
For attitudes, survey items capture agreement with statements regarding the respondent’s support for the provision of flexible working arrangements and/or crèches at the workplace. Higher values correspond to more favorable attitudes toward these other family-friendly workplace policies.
For beliefs about other family-friendly workplace policies, we will measure respondents’ expectations about the effects of flexible working arrangements and/or crèches at the workplace on teachers’ absenteeism, motivation or performance, fertility, and willingness to work in such schools.
3. Work satisfaction
We measure teachers’ satisfaction with their work, whether they would recommend the workplace as a great place to work, how satisfied they are with the facilities at the workplace, and how long they plan to work in the school.
4. Relationship with colleagues
This outcome family captures teachers’ perceptions of and satisfaction with workplace relationships and collegiality within the school environment. Survey items measure overall satisfaction with relationships among colleagues, trust in coworkers and school management, perceptions of interpersonal tensions or disagreements within the school, and expectations of mutual support among colleagues during periods of personal or family-related need. Higher values correspond to more positive workplace relationships, greater trust, and stronger perceived collegial support.
5. Well-being
We ask about life satisfaction in general.
6. Breastfeeding
In the endline, we ask teachers with children born in the past two years to report a detailed history of their breastfeeding behavior, at home and at the workplace, whether it corresponded to their preferences, and whether they had any negative experiences at work. We also have more general questions for all teachers about how supportive of breastfeeding they think their place of employment is and whether mothers breastfeed or express milk during working hours.
7. Work behavior
Our endline survey will also be used to look at some proxies for outcomes related to work behavior, including absences, working hours, productivity, and turnover. However, we aim to collect administrative data on these outcomes for more precise estimates.
Here are the ways in which we will use the surveys to measure work behavior:
Working time: We ask teachers to report how many days and hours they work each week.
Absenteeism: We ask teachers to report their absence days in the last six months, as well as absence days in the first three months after coming back from maternity leave or after having a child, for respondents to whom this applies. We also have school-level absenteeism proxies reported by the head teacher, for the school as a whole.
Productivity: We ask an attention scale, an adapted version of the Adult ADHD Self-Report Scale (ASRS), which allows us to measure self-reported attention in a work setting in the last three months or, for those who came back from leave, their attention after coming back. This is a self-reported proxy of productivity. We also ask female teachers with children younger than two to read a short text and correct any typos or grammatical mistakes. We will measure productivity as the share of typos correctly found.
Turnover: We will compare the list of teachers present at school at baseline and endline, according to our listing data. We will also ask the head teacher for the number of teachers who left or joined the school in the past year. This comparison and these questions will give us noisy measures of turnover at the school level, and potentially by gender.
For each family of outcomes, we will construct standardized indices following Kling et al. (2007). Specifically, each survey item will be coded so that higher values correspond to better outcomes or greater support for breastfeeding-related policies; items will be standardized using the control-group mean and standard deviation; and the index will be computed as the unweighted average of standardized items within each outcome family.
If some index components are missing, the index will be computed using the average of available non-missing components, provided at least half of the components are observed.
Heterogeneity of treatment effects: among other heterogeneity exercises, we will look at heterogeneity by distance to childcare centers, by school management practices, and by the presence of electricity and other infrastructure at school at baseline.