# Individual Characteristics and National Income Modify the Association between Cognitive Social Capital and Food Insecurity: Evidence from the Gallup World Poll, 2014–2021

## TL;DR

This study shows that social trust and support can reduce food insecurity, especially for vulnerable people in high-income countries.

## Contribution

The study reveals how social capital buffers food insecurity differently based on individual traits and national income levels.

## Key findings

- Higher social capital is consistently linked to lower food insecurity across all countries.
- The effect of social capital is strongest in high-income countries and among people with low income, poor health, or unemployment.
- Social capital reduces food insecurity by up to 20 percentage points for vulnerable groups.

## Abstract

Food insecurity affects >2.4 billion people globally and persists across income levels. Characteristics such as low income, low education, and unstable employment do not fully explain this persistence. Social resources embedded in networks, specifically cognitive social capital involving trust, reciprocity, and support, may offset constraints and buffer characteristics associated with a higher probability of food insecurity.

This study examined whether social capital was associated with a lower probability of food insecurity and whether this association varied by individual characteristics and country contexts, consistent with buffering and compensation.

Data were drawn from the Gallup World Poll (2014–2021), comprising 702,850 respondents aged ≥15 y across 115 countries. Moderate or severe food insecurity was assessed using the 8-item Food Insecurity Experience Scale; social capital was measured using a binary indicator. Six individual-level characteristics were tested using multilevel linear probability models with country fixed effects and interaction terms. Country-specific slope coefficients capturing the association between social capital and food insecurity, obtained from a random-coefficient model, were regressed on log-transformed gross national income (GNI) per capita.

The mean probability of food insecurity was 0.259; the mean prevalence of social capital was 0.820. In all countries, higher social capital was associated with a lower probability of food insecurity (slopes −0.318 to −0.055); quadratic analysis of the slopes on log-transformed GNI per capita showed consistent slopes in low- and middle-income countries and steeper slopes in high-income countries (P < 0.001). Among individuals, associations were largest with primary education (−16.96 pp), low income, unemployment (−20.4 pp), poor health (−18.44 pp), and widowhood (−19.68 pp).

The strength of the negative association between social capital and food insecurity varied by individual characteristics and national income, consistent with buffering and compensation.