Cash or Meals? Household Preferences in the Ghana School Feeding Programme

School meals are popular among governments and charitable organisations aiming to alleviate hunger in schools. Yet, there is often debate around whether alternative approaches, such as cash transfers or take-home rations, might achieve similar goals more effectively. Common questions include Wouldn’t cash transfers boost school enrolment and reduce hunger at a lower cost? and Could take-home rations relieve schools of the burden of delivering meals? These questions reflect the broader discussion of cash versus in-kind assistance in social safety net programs.

While there is growing academic and policy interest in the effectiveness of school meals versus cash grants, the opinions of households, especially parents and caregivers responsible for learners and their nutrition, are often overlooked. To fill this gap, we studied parental preferences for alternative modalities in the context of Ghana’s national school feeding programme. Our study involved a nationally representative sample of 1,425 caregivers whose children attend 373 primary schools participating in the programme. We asked parents if they would prefer take-home rations or cash grants over the on-site meals their children currently receive. The paper explores the demographic and socioeconomic factors influencing these preferences.

Caregiver preference highlights the influence of the existing setup and regional differences in perceived programme performance

A large majority of caregivers (around 77 percent) preferred on-site school meals to take-home rations or cash grants. For many, this choice may reflect a bias toward the current setup, as most children are already accustomed to on-site meals. But there is substantial concentration of caregivers wishing to deviate from the status quo, i.e. on-site meals, in certain regions. This is likely driven by regional variations in the performance of the programme. There is an 86 percent correlation between the proportion of parents in a region who prefer either of the alternative modalities to on-site meals (see Figure 1). Parents in metropolitan areas are slightly more likely to prefer take-home rations over on-site meals than those who live in more rural areas.

Figure 1. Higher preference for alternative modalities is concentrated in certain regions

Since all caregivers in our study have children receiving school meals, their views on alternatives often reflect their experience with the current programme. We asked caregivers to rate the programme based on nutritional quality, taste, and portion size, and created a composite index to assess programme performance. Caregivers who were dissatisfied with the current system were more likely to favour take-home rations or cash grants.

Caregivers consider several factors, including time constraints and local economic conditions, when choosing a preferred school meals modality

In our analysis of potential predictors of parental preference, we accounted for the composite performance index to eliminate the influence of current experience or status quo bias. Our findings show that caregivers consider multiple factors when selecting the most suitable modality for themselves and their children.

Mothers in more professional jobs tend to prefer on-site meals

A key consideration behind the choice of modality might be the time required to prepare meals at home in the case of cash grant or take-home ration. We found that mothers in professional occupations are 15 percent more likely to prefer on-site meals over take-home rations. While there is some evidence that professional mothers also preferred on-site meals to cash grants, these results are less robust. Generally, female caregivers tend to favour cash more than their male counterparts, which aligns with their role in managing household budgets and the flexibility cash grants offer. So, the preference of professional mothers for on-site meals could be interpreted as reflecting the effect of the opportunity cost of time.

Food price inflation is negatively associated with preference for cash grants

A key insight from the literature on cash vs in-kind food transfer is that the effectiveness of either modality depends on local market conditions. This is particularly relevant for our study because the survey was conducted during a period of high inflation and food price volatility in Ghana. Our results confirm the importance of local market conditions as parents residing in regions with relatively higher food prices prefer on-site meals to cash. But our findings also show that higher regional food prices are positively correlated with more parents preferring on-site meals to take-home ration. This is likely because caregivers would need to purchase complementary food items to convert take-home rations into meals.

We also used the caregiver’s own assessment of how their household had been impacted by inflation to further examine variations in modality preference. As expected, households significantly impacted by inflation were less likely to prefer cash grants. But, in this case, there seems to be no distinction between on-site meals and take-home ration in terms of the preferences of households impacted by inflation, reinforcing the cash vs in-kind argument.

Giving parents a voice is not just about agency, but also efficiency

While it may not be feasible to offer the choice of on-site meals, take-home rations, or cash grants simultaneously, school feeding programmes could be fine-tuned to address regional food price variations and household needs. This means implementing organisations could use a more differentiated mix of modalities to target different socioeconomic contexts, geographic regions, or even seasons of the year. Additionally, considering caregivers’ preferences not only recognizes their agency but can also improve programme efficiency and optimise outcomes for parents and students alike.