Determinants of food insecurity among households with children in Villa el Salvador, Lima, Peru: the role of gender and employment, a cross-sectional study

More than 820 million people in the world are hungry and approximately 2 billion are moderately or severely food insecure [1]. Food insecurity is the inability to obtain nutritious and sufficient food due to factors such as lack of monetary resources, food access, and food availability, increasing the risk of malnutrition and poor health [2]. Estimates of food insecurity reveal an upward trend in the number of people suffering from severe food insecurity worldwide, now magnified with the poverty exacerbation, food access decline, and mobility restrictions– which have limited access to work and to formal and informal food assistance resources – associated with the COVID-19 pandemic [1‐3].

From 2014–2019, Latin America and the Caribbean had the fastest growth of moderate to severe food insecurity compared to any other region worldwide, rising from 22.9% to 31.7% [1]. Given this rise, exploring factors related to moderate to severe food insecurity in Latin America is warranted. While the prevalence of food insecurity is slightly higher among women than men in every continent, Latin America has the largest food insecurity gender gap, with 32.4% of women and 25.7% of men being moderately to severely food insecure in 2019 [4]. Factors contributing to this gender inequity include underrepresentation of women in formal employment, heightened burden of the care for dependents on women, and unequal compensation of labor for women vs. men [1‐8]. These limitations for women may influence household food security status, where woman-headed households may be more vulnerable to be food insecure. For example, woman-headed households are more likely to live in poverty than man-headed households in Latin America [6].

There is a limited number of studies assessing the impact of the gender of the head of the household on food insecurity in Latin America; most existing studies have been conducted in the United States, Africa, and Asia [7‐10]. One study conducted in Brazil reported that woman-headed households were 32% more likely to be food insecure than man-headed households [11]. However, this study did not use a validated tool to assess food insecurity. A few studies have investigated household food insecurity and gender, using a previously validated tool, in Colombia, Honduras, and Brazil [12‐14]; these studies suggest that employment might be a modifying factor in the burden of food insecurity on woman-headed households. Given the high burden of food insecurity in Latin America and the Caribbean, and the potential role of inequitable employment status among women, the objective of this study was to explore the association between the gender of the head of the household, the employment status of household members, and food insecurity in a low-income district of Lima, Peru.

Of the 329 households surveyed, 254 (77.2%) were classified as food insecure, with 49 (19.3%) identified as mildly food insecure, 79 (31.1%) as moderately food insecure, and 126 (49.6%) as severely food insecure. Therefore, the minority group was the food secure, which represented only 22.8% (n = 75) of the sample. Demographic household characteristics are presented in Table 1.

The proportion of women-headed households increased with increased food insecurity severity. The mean age of the head of the household also increased with severity of food insecurity (mean age in years [SD]: food secure = 39.3 [12.4], mildly food insecure = 37.0 [11.2], moderately food insecure = 37.4 [12.3], severely food insecure = 44.2 [13.1]), while weekly household food expenditures per person decreased with increased food insecurity severity (mean in Peruvian soles [SD]: food secure = 43.0 [20.0], mildly food insecure = 37.9 [14.9], moderately food insecure = 36.4 [13.5], severely food insecure = 30.8 [14.4]). Regardless of food insecurity severity, the weekly household food expenditure per person in the sample was lower than the cost of the basic food basket in Peru, which was 46 soles in 2019; the basic food basket is defined as one including the most inexpensive foods a person or household can buy to meet the minimum daily food consumption to satisfy nutritional requirements [21]. Among food insecure household heads, 34% had less than high school educational attainment, 36% completed high school, and 30% had at least some college or technical school. More than two-thirds of food insecure household heads were not formally employed, 22% were self-employed, and only 11% were formally employed. The chi-square test of independence (p-value = 0.0207) provided evidence that gender of household head and household food insecurity were significantly associated.

Table 2 displays the results of the multivariable logistic regression models. In Model 1, woman-headed households had 2.8 times the odds of being food insecure compared to man-headed households, adjusting for age, education, and employment of the household head, and household food expenditure per person. Education had a significant association with food insecurity: compared to households with a household head who had some post-secondary education, households with household heads who did not complete high school and households with household heads with a high school education were 3.4 times and 2.3 times more likely, respectively, to be food insecure. Employment of the household head was not associated with food insecurity.

Similarly, in Model 2 (Table 2), woman-headed households had 2.6 times the odds of being food insecure compared to man-headed households, adjusting for age and education of the household head, household-level employment, and household food expenditure per person. Education was again significantly associated with food insecurity and household-level employment status was not associated with food insecurity.

In this study based on households with children living in a low-income district in Lima, Peru, we found that gender of the household head – defined as the person in charge of food purchases – is an important determinant of household food insecurity. Similar to what previous studies in Colombia, Honduras, and Brazil have reported [12‐14], we found that woman-headed households are at higher risk of food insecurity. Woman-headed households in our sample had a higher proportion of members not formally employed or self-employed in their homes; however, we found no association between employment status of the household head and household food insecurity. Gender disparities in type of employment were found to be associated with money shortage, a proxy indicator for food insecurity, in a study using data from Ecuador, Bolivia, Thailand, and the Philippines, with women in households with money shortages being more likely to be self-employed, as opposed to being salaried employees, when compared to women in households without money shortages [22]. This multi-country study used an indirect measure of food insecurity, contrary to our study which used HFIAS, a validated tool to measure food insecurity, which may explain the different results.

Given the disproportionate burden of food insecurity that falls on woman-headed households, strategies to overcome food insecurity should aim to 1) combat the social determinants that make women more vulnerable to food-specific poverty such as access to education, and 2) incorporate interventions aimed to influence gender dynamics/roles into national food security policies, including implementing efforts to reduce their non-income producing workload within the household [22]. Regardless of employment status, men earn more than women, on average, largely because women are more often employed in the informal sector of the economy in Latin America [23‐25], furthering women’s disadvantage. Moreover, women are often responsible for completing domestic tasks, limiting the number of hours they can use to generate income [23, 25]. In Peru, specifically, men earn on average 45% more than women; 6% of this wage gap is explained by individual characteristics (e.g., differential educational attainments), 11% by differences in supports (i.e., individual characteristics of men and women being distributed differently impeding appropriate matching), and 28% by non-observable individual characteristics, among which we can include social roles [26]. The burden women carry – through ascribed social norms regarding the type of work they do and household and childcare responsibilities – can reduce the income generating opportunities for woman-headed households [27, 28]. Previous interventions that have positively influenced gender dynamics within the household include in-person and video outreach to husbands and mothers-in-law, and women's support groups on financial and gender issues [29, 30].

This study has strengths and limitations. Strengths include cluster random sampling of households, the exploration of the impact of not only household headship but household composition, and the use of a validated tool to measure the outcome of food insecurity. Additionally, by defining household headship as the individual who oversees the household’s food purchasing, we were able to explore the importance and impact of gendered tasks that do not generate income. As for limitations, volunteer bias could have affected our results given our low response rate and generalizability of our results might be limited to other similar urban poor areas of Peru. Furthermore, it is possible that men answered our household questionnaire differently than women – especially the food insecurity questions – which may partly explain our results. Men might present higher social desirability bias while responding, as in a patriarchal society it might seem especially socially undesirable to be perceived to have failed to provide food for one’s family as a man [31]. Additionally, the number of man-headed households in our sample is significantly smaller than that of woman-headed households, affecting their comparability in terms of distribution by food insecurity severity.

Having a man-headed household in terms of overseeing food shopping could also be an indicator of household dynamics, such as a more liberal household in terms of not following gender roles. For example, a study using data from Mexico, Peru, and Ecuador found that countries with more biased gender social norms had women spend more time in unpaid tasks such as food shopping [32]. Previous studies have also found an association between women’s decision-making ability over household purchases (including food) and their age, employment, and number of dependents in households, all factors that have also been associated with household food insecurity [33]. While our screening question prior to initiating the household questionnaire explicitly asked for the person “in charge” of food purchasing, we cannot conclude to what extent those decisions are autonomous or driven by other household members. Given the high prevalence of informal employment in the study area, we did not ask participants about household income. While we controlled for food expenditures, the same expenditure amount could have very different implications for households at higher vs. lower income levels. Finally, our questionnaire did not allow to compare those without formal employment that are actively seeking and those that are not actively seeking this type of employment.

Food security is an intersectional issue related to income and poverty, access to productive resources, and gender-based discrimination [33]. Therefore, to ensure the reduction of food insecurity it is essential to integrate a gender equity perspective at all stages and levels of policies, programs and projects aiming to combat food insecurity paying particular attention to the social determinants of health (e.g., educational attainment) that intersect with gender [33, 34]. While we recognize this is a long-term effort, less resource- and time-intensive efforts to influence gender norms with the aim to decrease the unique burden on women to do unpaid care and domestic work can be implemented. These projects to influence gender norms must use interactive approaches to foster reflection, messages should be shared regularly and through a range of channels (e.g., social media, TV, radio, etc.), members and role models from the community must be called-in to participate in the program, and the program must engage men, while recognizing that big changes happen in small steps [35].

Women-headed households were found to be at higher risk of food insecurity compared to men-headed households in a low-income district in Lima, Peru. The gender dynamics affecting opportunities for remunerated employment and education for women should inform national food security policies and interventions with the goal of not only lowering food insecurity in the country, but also reducing gender inequities in food insecurity and other nutritional outcomes. Simultaneously, efforts should be placed in interventions addressing gender norms with the goal of reducing gendered workloads within the household and increasing women’s engagement in income producing activities.